HANDBOOK WRITING SKILLS

ON COMMUNICATING RESEARCH; WRITING & PRESENTING

STUFF

Remco Havermans, Caroline van Heugten, Katrijn Houben, & Robert van Doorn

2

Contents

3 An introduction to writing an empirical research paper

11 Referencing, fraud, and plagiarism

15 Overall article structure

18 An introduction to writing a literature review

35 Overall article structure

38 Style of writing

40 Formatting

41 Presenting research in PowerPoint

43 Overview of assignments in the FPN curriculum

44 Skills II

45 Skills III

54 Skills IV

55 Bachelor thesis

57 Master thesis

3

An introduction to writing an empirical researchpaper

When a scientist completes a study, s/he should communicate the results or else thestudy might just as well not have been conducted. Remember, a falling tree does notmake any sound if there is no one to hear it. In order to communicate the researchfindings, the researcher can write an article and submit it to a peer reviewed sciencejournal to be considered for publication, or s/he presents the study at some scienceconference. Usually, the researcher will do both. Writing and presenting are skills thatare very useful to the scientist (obviously), but are valuable in other professions aswell. More importantly, these are skills you can learn. The present guide concernswriting your very first research paper.

Research papers always have the same basic shape. The narrative structure of it issimilar to any good story: it has a beginning, middle, and an end.

In the beginning of your paper, you outline a (more or less) broad problem, fromwhich you argue to a more narrow and specific research question. Sketching a prob-lem is very important, because it provides the reader of your paper with a rationale foryour research question other than, for example: “I really like pussycats (I have three!)

BEGINNING(What’s the problem? What question

will be answered?)

MIDDLE(What did you do andwhat did you find inanswering the ques-

tion?)

END(Discuss your research findings and

your research methods. Draw conclu-sions regarding your original research

question and identify new ques-tions/problems)

4

and that is why I wanted to learn more about their behaviour…”. Think about it. Whoreads your papers? Right, that would be your teachers and supervisors. Trust me; theyare not at all interested in your personal life and preferences (unless you happen to beMichael Jackson-like famous). Therefore, try to convince your supervisor that you aresuper famous, or (and this is the preferred option) convince him/her that your researchquestion is a question that has general interest, for example: “We know a lot about thecharacteristic behavioural response patterns of dogs under a wide range of circum-stances. Surprisingly little is known about the behaviours of domestic cats, other thanthat they typically do not behave like dogs…”.

Paul Rozin (2009) recently agued that psychological science is about research thatexpands our understanding of how humans (and animals) behave, think and feel “andhow these events influence and are influenced by their material and social environ-ment” (p. 435). He listed a number of typologies of interesting papers from which I’vemade a selection:

1. ‘‘Here’s something interesting that no one has noticed, and it is not easily sus-ceptible to explanation by the principles available to us.’’

2. ‘‘Here’s something we haven’t studied, but it looks like it can be subsumedunder something we already know.’’

3. ‘‘Hey, someone did this really interesting study decades ago, and no oneseems to have noticed it.’’

4. ‘‘Everyone assumes Effect X, but is X robust and generalizable?’’

The beginning of a research paper is commonly referred to as the Introductionsection. As an example, see how I (together with a few select colleagues) wrote andorganized the ‘Introduction’ of a research paper I recently submitted for publication. Istart the Introduction with arguing that obesity is a problem (very broad), and thatdieting is hard. I argue that dieting is difficult because people want to experience thepleasure often associated with junk and snack food (like chocolate). Perhaps obesepeople find these foods more pleasurable/desirable than other people do. Here’s thespecific research question I formulate at the end of the Introduction: “… it remains tobe determined whether overweight/obese individuals … display this excessive ap-proach motivation for food and if so, whether this is … limited to high calorie snackfoods.” Note that one typically does not provide the introduction the header Introduc-tion.

The prevalence of overweight/obesity continues to rise world-wide and whether or not this should be termed an epidemic (oreven pandemic), severe obesity poses a serious health risk (seee.g., Olshansky et al., 2005). Obesity results from a positive en-ergy balance; energy intake exceeding energy expenditure

5

(Jéquier, 2002). At least from a health perspective, maintainingor losing weight is important and one of the most popular meth-ods for weight regulation is dietary restraint. Though popular itis notoriously hard to maintain a low calorie diet and thus anyachieved weight loss is difficult to uphold on the longer term(Wing et al., 2008). Indeed, dieting attempts appear associatedwith future weight gain, not weight loss (Lowe et al., 2006;Stice, Agras, & Hammer, 1999; Stice, Cameron, Killen, Hay-ward, & Taylor, 1999).Why does dieting so often fail? According to Lowe and Levine(2005), food intake is governed by physiological needs and he-donic eating motives. In most western societies where food isabundant, daily eating behaviour is likely determined by hedonicmotivations, that is, reward-driven eating or the desire to eathighly palatable foods. Due to the interplay between an innatesweet/fat preference and the rapid acquisition of strong prefer-ence for especially flavours that signal energy (Havermans &Jansen, in press), the most palatable foods for most people arevarious snack foods. In other words, eating is reward driven andthe foods that have the highest reward value are typically highcalorie snack foods.Hedonic eating encourages overeating and hence promotesweight gain. Clearly though, the hedonic appeal of a high caloriesnack is not equal for all individuals. For example, both re-strained eaters and overweight/obese persons, relative to con-trols, are more prone to work for high calorie snacks (Giesen,Havermans, Douven et al., 2010; Giesen, Havermans, & Jansen,2010; Giesen, Havermans, Nederkoorn, Strafaci, & Jansen,2009). Successful weight loss or the prevention of excessiveweight gain is often thought to hinge on the ability to suppress ahedonic eating motive (see e.g., Appelhans, 2009; Stroebe,Papies, & Aarts, 2008). Conceivably, this is all the more difficultif this appetitive motivation for palatable high calorie food isparticularly strong. According to this line of reasoning onewould expect that overweight/obese individuals show an exag-gerated approach motivation for especially high calorie foods.Indeed, Brignell, Griffiths, Bradley and Mogg (2009), using astimulus-response compatibility (SRC) task, demonstrated that‘overeaters’ (as based on their score on the ‘external eating’scale of the Dutch Eating Behavior Questionnaire) show a rela-tively exaggerated approach bias toward food cues.The SRC task was originally designed by de Houwer, Crombez,Baeyens, and Hermans (2001). Participants have to direct amanikin toward or away from a stimulus presented in the centreof a computer screen depending on given instructions. The sup-position is that one is faster to approach liked stimuli and toavoid disliked stimuli than one is to approach disliked stimuliand to avoid liked stimuli (Krieglmeyer, Deutsch, de Houwer, &de Raedt, 2010). In the SRC task employed by Brignell and col-leagues (2009), participants had to move the manikin either to-

6

ward or away from pictures related to either food or somethingother than food. The high-external eaters, relative to the low-ex-ternal eaters, were significantly faster to move the manikin to-ward the food pictures (and away from the non-food relatedcues) than they were to move the manikin away from the foodcues (and toward the non-foods) (see also Veenstra & De Jong,in press). Although interesting, it remains to be determinedwhether overweight/obese individuals too display this excessiveapproach motivation for food and if so, whether this is more orless limited to high calorie snack foods. For the present study,we hypothesized that overweight/obese persons display astronger approach bias toward specifically high calorie foodsthan normal weight controls do. To test this hypothesis we em-ployed an SRC task very similar to the task used previously byBrignell et al. (2009).

The middle part of the research paper comprises two larger sections: Methodand Results. The Method section describes what you did in trying to answer your re-search question. It always comprises the same subsections. Firstly, you include a sub-section “Participants”. Here you describe the sample of participants in your research.How many people did you observe or test? How many men and how many womendid you include in your sample? Were they of any specific age range? Provide all therelevant participant variables. For example:

Method

ParticipantsEighty-eight undergraduate students from Maastricht Universitywere invited to take part in a study on choice behaviour andmental fatigue. They in fact took part in two separate experi-ments: the present study and another study using a behaviouraltask to measure the relative reinforcing value of snack food, re-sults of which are reported elsewhere (see Giesen et al., 2010).Students were invited to take part on the basis of self-reportedBody Mass Index (BMI; kg/m2). We strived to create twoweight-discordant groups by selecting candidate participantswith a self-reported BMI 25 and an equal number of studentswith a self-reported BMI < 25 (preferably around 20). Partici-pants’ BMI was measured at the end of participation and the fi-nal assignment of participants to either the overweight/obesecondition or the normal weight condition was determined on thebasis of these actual BMIs. Participants with a BMI 25 wereclassified as overweight or obese, participants with a BMI be-tween 18 and 25 were classified as normal weight.

7

Secondly, a subsection Materials is intended to describe the materials and equipmentyou used to conduct the study with. For example:

MaterialsHunger ratings. Participants rated their momentary hunger on a100 mm Visual Analogue Scale (VAS) ranging from 0 “not at allhungry” (left anchor) to 100 “very hungry” (right anchor).Stimulus response compatibility (SRC) task. The task the par-ticipants had to perform was adapted from the SRC task de-scribed by Thewissen, Havermans, Geschwind, van den Hout,and Jansen (2007) and was programmed in E-prime (PsychologySoftware Tools, Inc). The task comprised two blocks of 80 trials.In each trial, either a 160 mm high x 215 mm wide food picture(10 different food pictures were used; e.g., a picture of chocolatecookies, grapes, crisps, et cetera) or a neutral picture (10 differ-ent neutral non-food pictures were used; e.g., a wooden shoe, atraffic cone, a cardboard box, et cetera) was displayed at the cen-tre of the screen together with a manikin (an approximately 18mm high x 10 mm wide matchstick figure) placed exactly be-tween the outer border (either the upper or lower border) of thepicture and the edge of the screen.Each block of trials had a different stimulus response assign-ment. In one block the participants were instructed to approachthe food pictures with the manikin by using the up- or down ar-row keys depending on the starting position of the manikin, andto avoid the neutral picture by moving the manikin away fromthe picture outside the screen. The following trial was initializedupon completion of the correct response (i.e., approach or avoid-ance) and the time to complete each trial was recorded (in ms).A correct response required a minimum of five consecutive keypresses. In the other block, participants received the instructionto avoid the food pictures and to approach the neutral pictures asquick as possible. The order of these two blocks and the eightdifferent trial types (position [above or below] x picture [foodvs. neutral] x instruction [approach vs. avoid]) per block was de-termined randomly for each separate participant.The task was preceded by a practice session that comprised twoblocks of 16 trials each. In one block they had to approach pic-tures depicting a chair and to avoid pictures displaying a lamp.This stimulus response assignment was reversed for the otherblock.Three Factor Eating Questionnaire. We employed a Dutchtranslation of the Three Factor Eating Questionnaire (TFEQ;Stunkard & Messick, 1985) comprising 51 items pertaining todietary restraint (i.e., the intention to control body weight by re-stricting one’s food intake), disinhibition, or hunger susceptibil-ity.

8

Thirdly, the Method includes a description of your study procedure in which you de-scribe in detail how you conducted your research. The reader of your paper should beable to exactly replicate the procedure of your study on the basis of the information inthe subsection Procedure. For example:

ProcedureThe experiment was approved by a local ethical committee. Par-ticipants were instructed to eat two hours prior to the experimentand from that time on to refrain from food until participation.Participants were tested individually between noon and 6 PM.The experiment started with a brief verbal description of thegeneral procedure of the experiment. After this, all participantsalso received written information regarding the experiment andwere asked to sign a consent form if they still wished to partici-pate.The participant was asked to rate his degree of momentary hun-ger on a VAS, after which he first completed a task designed toassess the reinforcing value of a given snack food relative to thevalue of a fruit or vegetable alternative. Participants had to workfor their highest liked snack food and their highest liked lowcalorie food (i.e. fruit or vegetable). Throughout the task thework effort required for snack food was gradually increased,thus allowing us to determine the relative motivation to work foror acquire high calorie snack foods. The task procedure is de-scribed in detail by Giesen, et al. (2010). This task was generallycompleted within 20 min and note that any ‘earned’ food wasconsumed only after the completion of both this so termed con-current schedules task and the SRC task.Upon completion of the SRC task, the participant was furtherasked to fill out the TFEQ. At the end of the study, the partici-pant’s height and weight was measured to calculate actual BMI.The participant was then thanked and received either a coursecredit or a € 7.50 monetary voucher as remuneration.

In case of more complex studies, the author usually includes a subsection termed DataReduction & Analyses describing how the data were collected, processed and ana-lyzed. For example:

Data Reduction and AnalysesAll analyses concerned the participants’ response time latenciesto the different trial types and stimulus response assignments inthe SRC task. To remove outliers, reaction times were excludedper participant and per trial type if they were smaller than 200ms or larger than 2.5 SD above the mean RT (10% of the data)(see also Thewissen et al., 2008).Weight, gender, and food approach/avoidance. To assesswhether overweight/obese participants show a stronger food ap-proach bias than the normal weight participants do, response

9

times for each participant (food approach vs. food avoidance)were averaged for the food approach assignment and the foodavoidance assignment. These averaged RTs then served as thedependent variable in a 2 (Group: overweight/obese vs. normal-weight) x 2 (Gender: male vs. female) x 2 (Assignment: foodapproach vs. food avoidance) ANCOVA with hunger ratings asa covariate. Hunger ratings were entered as a covariate becausehunger likely plays a role in food approach-avoidance motiva-tion and the degree of self reported hunger seemed to differ atleast marginally significant between the overweight/obese (MVAS hunger rating = 49.6) and the normal weight participants (M VAS

hunger rating = 57.5) (see also Table 1). Likewise, we added Genderas a factor to the analyses to control for any potential gender dif-ferences in food approach-avoidance motivation.

The Results section does not have a fixed number of subsections. You can use subsec-tions if you feel that provides a better read of the results. The general idea however, isto provide an overview of the results of your statistical analyses. Describe the statisti-cal tests and the relevant test values (e.g., t in case of a t-test, F in case of an F testwith an analysis of variance). Provide tables and/or figures describing and depictingmean scores of the relevant dependent variables for the different levels of your inde-pendent variables. Tables and figures are numbered. The table legend is presented atthe top of the table. The figure legend, describing exactly what one can see in the fig-ure, is presented just below the figure. For example:

Results

1200

1225

1250

1275

1300

1325

1350

1375

1400

1425

lean overweight lean overweight

male female

mea

n R

T (m

s)

food approachfood avoidance

Figure 1. Mean RT (+SEM) per response assignment per genderand for each group.

In Figure 1, the mean RTs per assignment are displayed for eachgroup and separately for each gender. When comparing the foodavoidance assignment RTs and the food approach assignment

10

RTs, no main effects were found (largest F[1, 83] = 2.75, p =.10), but a near significant Group x Gender interaction wasfound, F(1, 83) = 3.75, p = .06, 2

partial = .04. This two-way in-teraction though is qualified by the overarching Group x Genderx Assignment interaction, F(1, 83) = 5.09, p = .027, 2

partial =.06. In post-hoc analyses, we examined a potential two-wayGroup x Gender interaction for the food approach and foodavoidance assignment separately. This interaction was foundonly for the food avoidance assignment RTs, F(1, 83) = 5.16, p= .026, 2

partial = .06. We further examined this interaction bytesting for a difference in mean RTs between the over-weight/obese and normal weight participants for each genderseparately by means of independent samples t-tests. These testsrevealed a significant difference between the overweight/obesemen and normal weight men (t[42] = 2.24, p = .03), but not be-tween the overweight/obese women and the normal weightwomen in the sample (t[42] = 1.03, p = .31). Overweight/obesemale participants were much slower to avoid food stimuli thanthe normal weight male participants were.

The end of the paper is the Discussion section in which you compare the resultswith what you hypothesized at the end of the Introduction. Discuss the outcome of thestudy in a broader theoretical context. Identify new questions for further research.And do not hesitate to be critical of your own work. Discuss limitations of your studyand describe how to take these limitations into consideration in future research. Forexample:

Discussion

In the present study, it was investigated whether over-weight/obese persons, as compared to normal weight controls,show an exaggerated approach motivation for especially highcalorie foods. The pattern of results suggests that this is not quiteso. Rather, it appears that overweight/obese persons find itharder to avoid especially high calorie foods relative to normalweight individuals. One may speculate that a strong hedonic eat-ing motive makes it very hard to ‘walk away’ from these foods.But note that this pattern of results was restricted to the maleparticipants.Considering the response tendencies for the high calorie foods,overweight/obese women – relative to the normal weight women– showed a weaker relative approach tendency. Apparently, theoverweight/obese women were ambivalent toward high caloriefoods, ambivalence expressed as a relatively small approach ten-dency. It is conceivable that dietary restraint underlies this latterfinding. When one tries to lose weight, high calorie foods areusually ‘forbidden’ and thus it makes sense that given the highTFEQ-restraint score especially the overweight/obese women

11

displayed a strong avoidance motive when presented with thesefoods.In sum, the present study, demonstrates clear gender differencesin approach/avoid motivation regarding high calorie foods. Menand women can have very different reasons for overeating. Inter-estingly, Drapeau et al. (2003) found that a high restraint scorein women is associated with weight gain in a 6 year prospectivestudy, but an opposite pattern (i.e. restraint associated withweight loss) was found in men. This finding is difficult to rhymewith the present results. Why would overweight/obese womenstill tend to overeat on high calorie foods if they demonstrate aclear avoidance intention toward these foods? Perhaps the over-eating only occurs in situations where one simply cannot avoidbeing exposed to palatable food cues. Indeed, such exposure hasbeen shown to induce craving and to promote eating in particu-larly overweight subjects (see e.g., Jansen et al., 2003). Ofcourse, this account results from post-hoc reasoning and requiresfurther research. It would be interesting to examine whether foodcue exposure decreases food approach-avoidance ambivalence.The current investigation highlights the broader notion that gen-der differences in eating behaviour need to be accounted for inany model of weight control and/or overeating. Therefore, welike to reiterate the observation and recommendation made byRolls et al. (1991; p. 134): “… most studies of eating behavior donot attempt to distinguish between the sexes. There are, how-ever, several investigations indicating that this should be a partof future work.”

Referencing, fraud, and plagiarism

A big difference between a novelist and a scientist is that the novelist can lie. Scien-tists cannot lie in their writings. If they would lie all the time, nobody would be ableto trust anything any scientist says or claims anymore. That’s why we – as scientists –take science fraud very seriously. Researchers can make all kinds of claims, but theseclaims need to be backed up by research. The author thus also needs to refer to thatresearch source so that the reader may consult the source to check whether the authorisn’t lying or bluffing. Furthermore, it is often tempting for students to simply paste apaper together with beautifully long sentences and even whole paragraphs cut fromthe ready written works of someone else. Do NOT do this, this is considered plagia-rism. Even when you refer to the original work, with quotes, you cannot write a papercomprising only quotes. You are the author; you need to do the writing, even if thatsometimes means paraphrasing a perfectly fine sentence!

The standard way of referring to sources within the science literature by psychologistsis outlined by the American Psychological Association (henceforth APA) in their

12

Publication Manual. The rules for within-text referencing are that you mention thesurnames of the authors of the consulted source and the year in which that study waspublished in parentheses. For example:

According to Lowe and Levine (2005), food intake is governed by physiological needs and hedoniceating motives.

Always mention the surname of the first author. If the study published was written bytwo authors, then you always mention both authors. You can refer to that study di-rectly as in the example above or indirectly as in the example below:

Food intake is governed by physiological needs and hedonic eating motives (Lowe & Levine, 2005).

Note that in this example, the author surnames are also placed between parenthesesand now the conjunction ‘and’ is replaced by ‘&’.

To complicate things further, APA rules dictate that if you refer to an article or bookchapter that has three, four, or five authors, you need to list the surnames of all theseauthors the first time you refer to this source. From then on, however, you only men-tion the surname of the first author followed by et al. (which means “et alii” which isLatin for “and others”) as in the following examples:

Direct reference:To test this hypothesis we employed an SRC task very similar to the task used previously by Brignell etal. (2009).Indirect reference:To test this hypothesis we employed an adapted version of an SRC task used in a previous study(Brignell, et al., 2009).

When the source publication you wish to refer to has more than 5 authors, you alwaysmention the surname of the first author followed by ‘et al.’ (as in the example above).

After the Discussion section, on a separate page, you include a list of all the refer-ences you used in your paper. Here’s the example:

References

Appelhans, B. M. (2009). Neurobehavioral inhibition of reward-driven feeding:Implications for dieting and obesity. Obesity, 17(4), 640-647.

Brignell, C., Griffiths, T., Bradley, B. P., & Mogg, K. (2009). Attentional and ap-proach biases for pictorial food cues: Influence of external eating. Appetite,52, 299-306.

De Houwer, J., Crombez, G., Baeyens, F., & Hermans, D. (2001). On the general-ity of the affective Simon effect. Cognition & Emotion, 15, 189-206.

13

Drapeau, V., Provencher, V., Lemieux, S., Després, J. P., Bouchard, C., &Tremblay, A. (2003). Do 6-y changes in eating behaviors predict changes inbody weight? Results from the Québec Family Study. International Journalof Obesity, 27(7), 808-814.

French, S. A., Jeffery, R. W., & Wing, R. R. (1994). Food intake and physical ac-tivity: A comparison of three measures of dieting. Addictive Behaviors, 19,401–409.

Giesen, J. C. A. H., Havermans, R. C., Douven, A., Tekelenburg, M., Nederkoorn,C., & Jansen, A. (2010). Will work for snack food: The association of BMIand the reinforcing value of snack food. Obesity, 18, 966-970.

Giesen, J. C. A. H., Havermans, R. C., & Jansen, A. (2010). Substituting snackswith strawberries and sudokus: Does restraint matter? Health Psychology,29(2), 222-226.

Giesen, J. C. A. H., Havermans, R. C., Nederkoorn, C., Strafaci, S., & Jansen, A.(2009). Working harder to obtain more snack foods when wanting to eat less.Behaviour Research and Therapy, 47(1), 13-17.

Havermans, R. C., & Jansen, A. (in press). Acquired tastes: Establishing food (dis-)likes by flavour-flavour learning. In V. R. Preedy, R. R. Watson, & C. R.Martin (Eds.), International Handbook of Behavior, Diet, and Nutrition. NewYork: Springer.

Jansen, A., Theunissen, N., Slechten, K., Nederkoorn, C., Boon, B., Mulkens, S.,& Roefs, A. (2003). Overweight children overeat after exposure to food cues.Eating Behaviors, 4, 197-209.

Jéquier, E. (2002). Pathways to obesity. International Journal of Obesity, 26(Suppl.), 12-17.

Krieglmeyer, R., Deutsch, R., De Houwer, J., & De Raedt, R. (2010). Beingmoved: Valence activates approach-avoidance behavior independently ofevaluation and approach-avoidance intentions. Psychological Science, 21(4),607-613.

Laessle, R. G., Tuschl, R. J., Kotthaus, B. C., & Pirke, K. M. (1989). A compari-son of the validity of three scales for the assessment of dietary restraint.Journal of Abnormal Psychology, 98, 504–507.

Lowe, M. R., Annunziato, R. A., Markowitz, J. T., Didie, E., Bellace, D. L., Rid-dell, L., Maille, C., McKinney, S., & Stice, E. (2006). Multiple types of diet-ing prospectively predict weight gain during the freshman year of college.Appetite, 47(1), 83-90.

Lowe, M. R., & Levine, A. S. (2005). Eating motives and the controversy over di-eting: Eating less than needed versus less than wanted. Obesity Research,13(5), 797-806.

Olshansky, S. J., Passaro, D. J., Hershow, R. C., Layden, J., Carnes, B. A., Brody,J., Hayflick, L., Butler, R. N., Allison, D. B., & Ludwig, D. S. (2005). A po-tential decline in life expectancy in the United States in the 21st century. NewEngland Journal of Medicine, 352(11), 1138-1145.

Rolls, B. J., Fedoroff, I. C., & Guthrie, J. F. (1991). Gender differences in eatingbehavior and body weight regulation. Health Psychology, 10(2), 133-142.

Stice, E., Agras, W. S., & Hammer, L. D. (1999). Risk factors for the emergenceof childhood eating disturbances: A five-year prospective study. Interna-tional Journal of Eating Disorders, 25(4), 375-387.

Stice, E., Cameron, R. P., Killen, J. D., Hayward, C., & Taylor, C. B. (1999).Naturalistic weight-reduction efforts prospectively predict growth in relative

14

weight and onset of obesity among female adolescents. Journal of Consultingand Clinical Psychology, 67(6), 967-974.

Stroebe, W., Papies, E. K., & Aarts, H. (2008). From homeostatic to hedonic theo-ries of eating: Self-regulatory failure in food-rich environments. Applied Psy-chology: An International Review, 57(Suppl 1), 172-193.

Stunkard, A. J., & Messick, S. (1985). The three-factor eating questionnaire tomeasure dietary restraint, disinhibition and hunger. Journal of PsychosomaticResearch, 29(1), 71-83.

Thewissen, R., Havermans, R. C., Geschwind, N., van den Hout, M., & Jansen, A.(2007). Pavlovian conditioning of an approach bias in low-dependent smok-ers. Psychopharmacology, 194(1), 33-39.

Van Gucht, D., Vansteenwegen, D., Van den Bergh, O., & Beckers, T. (2008).Conditioned craving cues elicit an automatic approach tendency. BehaviourResearch and Therapy, 46(10), 1160-1169.

Veenstra, E. M., & de Jong, P. J. (2010). Restrained eaters show enhanced auto-matic approach tendencies towards food. Appetite, 55(1), 30-36.

Wing, R., Papandatos, G., Fava, J. L., Gorin, A. A., Phelan, S., McCaffery, J., &Tate, D. F. (2008). Maintaining large weight losses: The role of behavioraland psychological factors. Journal of Consulting and Clinical Psychology,76(6), 1015-1021.

These references are placed in alphabetical order on the basis of the surname of thefirst author of the paper. Again, there are APA rules to follow. The Rules for format-ting your reference list can be found in the APA Publication Manual. There are manycopies of this manual at the Maastricht University Randwijck Library (Studieland-schap). You can also buy/order your own copy at the Study Store (the bookstore infront of the north entrance of the UNS40 building). There are also many websites ex-plaining the finer details of how to prepare your paper and endnotes in APA style.There is a neat APA style tutorial to be found at www.apastyle.org. In the currentmanual I’ll limit myself to providing a mere two examples of the more commonlyused sources: a journal article, and book section.

Journal reference:First type the names of the authors (surnames followed by initials). Next specify theyear of publication and place this between parentheses. Type the article title in sen-tence case and the journal title in title case. This means that if you refer to an articlepublished in the journal Health Psychology you do not type ‘Health psychology’ (thisis sentence case with only the first word starting with an upper case), but you type‘Health Psychology’. Italicize the journal title and volume number. Here’s the exam-ple:

Giesen, J. C. A. H., Havermans, R. C., & Jansen, A. (2010). Substituting snacks with strawberries andsudokus: Does restraint matter? Health Psychology, 29, 222-226.

Book section:

15

After the chapter title, type ‘In’, the editor’s name, the abbreviation ‘Ed.’ (in paren-theses) and then the title of the book. Give the page numbers in parentheses after thebook title. Example:

Havermans, R. C. (2009). Increasing children's liking and intake of vegetables through experientiallearning. In R. R. Watson & V. R. Preedy (Eds.), Bioactive Foods in Promoting Health: Fruitsand Vegetables (pp. 273-283). San Diego: Elsevier.

Overall article structure

The research paper introduction is usually preceded by an abstract, a concise sum-mary (no longer than 150 words) describing the whole study in a single paragraph.The abstract is important as it contains the information on the basis of which its readerwill decide to continue reading the whole article. That is why the abstract needs tocontain all the relevant information and at the same time be as brief as possible. Theabstract derives from the rest of the article and that’s why you need to write the ab-stract only after you’ve completed the rest of the article (Glasman-Deal, 2010). Donot insert any references or subheadings in your abstract. You may add keywords tothe abstract. For example:

Abstract

In this study, we hypothesized that overweight/obese personshave an exaggerated approach tendency toward high caloriefoods. Testing this hypothesis, a stimulus response compatibility(SRC) task was used to assess approach-avoidance tendenciestoward food in both overweight/obese participants (n = 42), andnormal weight controls (n = 46). The SRC task is a reaction timetask measuring how fast one approaches and avoids pictures offood and non-foods according to given instructions. It was foundthat overweight/obese men are slower at avoiding particularlyhigh calorie snack foods. But this does not appear to be the casefor overweight/obese women who showed nearly as fast avoid-ance as approach toward the high calorie food cues. It is con-cluded that overweight/obese women, rather than men, are am-bivalent toward high calorie foods, which is the likely result ofhigh dietary restraint.

Keywords: appetite; approach-avoidance; dietary restraint; gen-der; obesity

The title page precedes the abstract. It should contain the title (really!). The title can-not be longer than 15 words and needs to be informative. Hilary Glasman-Deal (2010;

16

p. 225) aptly notes: “The title should predict and describe the content of the paper asaccurately as possible”. Glasman-Deal is right, but I would like to add that the titleshould also attract readers. If no one decides to read your paper, you might as wellhave saved yourself the trouble of writing the paper in the first place. Remember thefalling tree? The original title of the paper serving as the example in this manual was:

“Individual differences regarding automatic approach-avoidance tendencies towards high energy densefoods”

This title covered the content of the paper, but probably would not attract a wide read-ership. We finally settled for the following, more pithy title:

“Weight, gender, and snack appeal”

On your student research paper you also need to provide an overall word count (howmany words does your paper count? – make Word do the counting for you) your fullname(s) and student ID, the name of your supervisor/mentor, and that you’re a studentat Maastricht University, Faculty of Psychology & Neuroscience.

Here then is the general structure of the paper:

17

INTRODUCTION (page3)

(What’s the problem? What questionwill be answered?)

DISCUSSION(Discuss your research findings and

your research methods. Draw conclu-sions regarding your original research

question and identify new ques-tions/problems)

METHOD(What did you do toanswer the research

question?)

RESULTS(What did you find?)

TITLE (page 1)

ABSTRACT (page 2)

REFERENCES (new page)

18

An introduction to writing a literature review

One of the most important types of articles published is the literature review. It is im-portant because a good theoretical review is often the starting point for further empiri-cal research (Fernandez-Rios, & Buela-Casal, 2009). There are in fact several formsof reviews including (1) a summarizing and exhaustive review providing a state of theart overview of a body of research on a given topic (i.e., “Women and their claims tomulti-tasking: a review of more than three decades of research”), or (2) an opinionpiece concerning an empirically justified argument for a given statement (i.e.“Women cannot do multiple different things at the same time”), or (3) a description ofa (somewhat) novel theoretical concept capable of explaining and integrating appar-ently discordant findings (i.e. “Cognitive biases and illusions, and females’ falseclaims to multi-tasking”). I’ll give examples of some of my own work (i.e.,Havermans & Jansen, 2007) concerning a blend of all three forms in this chapter.

Whatever its form, the primary functions and purpose of a review paper is toprovide the reader with a succinct and scholarly summary of the relevant researchliterature that is not just descriptive but also offers a novel insight and that informsand inspires further research.

It is often exclaimed that literature review papers are not as structured or asstraightforward as an empirical article. This is true, but only to a degree. As I stated inthe previous chapter on writing an empirical research article, any research paper al-ways has the same basic shape. This includes a literature review as such a review tooanswers a specific research question. Again, I can refer to the following figure outlin-ing the basic structure of your literature review: it has a beginning, middle, and anend.

A literature review is research based on your reading and understanding of therelevant empirical research literature to indentify relations, contradictions, gaps, andinconsistencies concerning a given topic. The topic of your paper can be anythingreally as psychology as a science is about everything and everyone. That doesn’tmake it easier to select a topic, but don’t think too soon that your topic of interest isunfit for a psychology review paper.

19

The beginning of your review should concern a description of your topic as thecentral problem of your paper. The introduction should culminate to a specific re-search question and aim. As a rule, this question should be as specific as a questionfor empirical experimental research. An example of a too boring and too broad ques-tion is: “What are the differences between men and women?” A better question wouldbe: “Are women really better at multi-tasking than men are and if so, why?” The latterquestion is much more specific and allows for the refutation or more nuanced view ofa conventional wisdom, which is always informative and fun. Remember that youstart your introduction with a broad description of the problem. But by the end of yourintroduction you should not just state what topic or specific question you will address,but also what the aim(s) and further structure of your paper is. As an example here’sthe introduction of a 2007 paper I wrote (with Anita Jansen) on evaluative condition-ing.

As preferences and aversions are important determinants of humanbehaviour, much research has been devoted to understanding thenature of human likes and dislikes. Some preferences and aver-sions are innate, such as the preference for sweet tastes and theaversion for bitter tastes. However, most preferences and aversionsare acquired with experience (Capaldi, 1996). Associative learninghas been proposed as the primary mechanism underlying such de-

BEGINNING(What’s the problem? What question

will be answered?)

MIDDLE(What did you do andwhat did you find inanswering the ques-

tion?)

END(Discuss your findings. Draw conclu-sions regarding your original questionand identify new questions/problems)

20

velopment of likes and dislikes (De Houwer, Thomas, & Baeyens,2001). In 1975, Levey and Martin demonstrated associatively ac-quired affect. In this seminal study they first let participants cate-gorize pictures of paintings as liked, neutral, or disliked. In thesubsequent conditioning phase, neutral pictures were consistentlypaired with either a liked, disliked, or another neutral picture. Attest, where participants had to evaluate the pictures that had beenpresented during conditioning, it was found that the neutral pic-tures paired with the disliked stimuli were evaluated as more nega-tive. Similarly, the neutral pictures paired with the liked pictureswere now evaluated as more positive. Martin and Levey (1978)coined the term evaluative conditioning (EC) in referring to thisapparent associative learning of preferences and aversions.

EC appears to be a reliable effect. Numerous picture-picture studies have successfully demonstrated transfer of affect,using pictures of fountains and sculptures (e.g., Hammerl & Gra-bitz, 1993), or pictures of human faces (e.g., Baeyens, Eelen, Vanden Bergh, & Crombez, 1989). Apart from being a reliable effect,EC also appears to be a general phenomenon. The transfer of af-fect has been shown to occur using different flavours as stimuli(e.g., Zellner, Rozin, Aron, & Kulish, 1983; Baeyens, Eelen, Vanden Bergh, & Crombez, 1990; Havermans & Jansen, 2007), orhaptic stimuli (Hammerl & Grabitz, 2000). Several studies havealso demonstrated cross-modal EC (see e.g., Todrank, Byrnes,Wrzesniewski, & Rozin, 1995). It should be noted though thatcross-modal EC appears to be somewhat less robust than EC usingstimuli of the same perceptual modality, as demonstrated by sev-eral failures to obtain cross-modal EC (see Rozin, Wrzesniewski,& Byrnes, 1998).

Although EC has been studied extensively over the lastthree decades there is still considerable debate concerning the as-sociative nature of the learning of likes and dislikes. In this chap-ter, we summarize different demonstrations of EC and discuss towhat degree these findings can be conceptualized as a form of Pav-lovian stimulus-stimulus learning. The foremost accounts of ECare evaluated in the light of this discussion. It is argued that noneof the present models of EC provide a satisfactory account of EC.In the second part of this article we discuss the development of anew model to account for the variety of EC effects.

As a literature review is also guided by a research question it makes sense to also in-clude (at least) a paragraph on your methods for finding and selecting papers. Here isan example from Abigail Remick and colleagues who included the following para-graph at the end of their introduction to their 2009 review paper on the ‘variety effect’published in Psychological Bulletin (p. 434):

“We located articles for this review through computerized searchesof the psychology database PsycINFO (through 2008) using combi-nations of the following keywords: variety, food, eat(ing), diet, sen-

21

sory-specific satiety, and monotony. We also located referencesthrough cross-references within articles and by using the “cited by”function in PsycINFO. Articles that were chosen for this review usedhuman participants and assessed various moderators of the varietyeffect.”

Discuss the usefulness of adding such a paragraph (or perhaps even a whole section)with your supervisor. You are allowed to omit this paragraph if your supervisor feelsit does not really fit your paper.

The middle part of your review concerns your arguments for or against yourposition. Note that these arguments need to be based on the available research on yourtopic. In other words, your arguments need to be empirically and epistemologicallyjustifiable. It is useful to not just summarize a lot of research in a single middle sec-tion. Try to structure it into two or three major sections and ensure that there is a logi-cal flow from one section to the next. In my review of evaluative conditioning I firstsummarize typical findings of evaluative conditioning and discuss to what degree thiscorresponds to classical conditioning and the next big section then is reserved for thepresentation of an alternative model explaining evaluative conditioning. Here’s thetext:

The associative nature of evaluative conditioningIn general, it is suggested that EC comprises the learning of an as-

sociation between a neutral stimulus and an affective stimulus. ECthus appears similar to Pavlovian stimulus-stimulus learning. In Pav-lovian conditioning (PC), one learns an association between a neutralconditioned stimulus (CS; e.g., a tone, or a light) and a biologicallyrelevant unconditioned stimulus (US; e.g., food, or an electric shock).Due to the formation of such an association, the mere presentation ofa CS comes to elicit conditioned responding in anticipation of the US(e.g., appetitive behaviour when the US is food, or fear-motivatedbehaviour when the US is an electric shock) (Pearce & Bouton,2001). In EC, the neutral stimulus is usually referred to as the CS andthe affective stimulus as the US. As De Houwer et al. (2001) note,there exist both similarities and discrepancies between findings in PCand EC.

Similarities between EC and PC. In PC an association between theCS and US gradually increases in strength to an asymptotic valuewith an increasing number of trials, which can be represented by atypical negatively accelerated learning curve. Some EC studies havefound such a significant gradual change in affective value of the CSwith an increasing number of trials (e.g., Baeyens, Eelen, Crombez,& Van den Bergh, 1992). However, De Houwer et al. (2001) notethat there have also been failures at demonstrating such a gradual ac-quisition of affect. This can be explained by arguing that the forma-tion of an evaluative association occurs exceptionally rapid. Indeed,

22

significant transfer of affect can be shown after just a single pairingof the CS and the affective US (Stuart, Shimp, & Engle, 1987).

There are also several functional characteristics of EC that parallelPC. One of these characteristics concerns the demonstration that ECis sensitive to a counterconditioning procedure. In demonstratingcounterconditioning of acquired affective value, Baeyens et al.(1989) conducted a picture-picture EC study. After baseline ratingsof the pictures, neutral CSs were paired with either positive, or nega-tive USs. After the acquisition phase the CSs were rated as morepositive (when paired with a liked picture) or more negative (whenpaired with a negative US), indicative of EC. After the post-acquisition assessment of likes and dislikes, some CSs were nowpaired with a US of opposite affective value as during the acquisitionphase. The other CSs were either presented alone or not presented atall. After this phase of the experiment participants had to rerate allthe CSs. The rating of the CSs that had been paired with the US ofopposite affective value returned to baseline level, whereas evalua-tion of the other CSs did not change in comparison with the post-acquisition ratings.Hammerl and Grabitz (1996) demonstrated in two experimentsevaluative sensory preconditioning, also using a picture-picture para-digm. In the preconditioning phase, participants received several se-quential pairings of two neutral CSs. In the second phase, CS2 waspaired with a liked US. At test, transfer of positive affect was notonly apparent for CS2 but also for CS1 that had not been paired withthe US. This finding corresponds well with the demonstration of Pav-lovian sensory preconditioning (see e.g., Rescorla & Durlach, 1981).As in PC, EC also appears sensitive to stimulus preexposure effects.For instance, Stuart et al. (1987) demonstrated that CS preexposureattenuates EC. More recently, De Houwer, Baeyens, Vansteenwegen,and Eelen (2000) also found some indication that preexposure to theCS may attenuate the transfer of affect to this CS in a subsequent ECprocedure. Similarly, EC is sensitive to US preexposure. Hammerl,Bloch, and Silverthorn (1997) found that repeated preexposure of theUS in a picture-picture study attenuated EC. This is in accordancewith the finding that US preexposure leads to reduced conditionedresponding in PC (see Randich & Lolordo, 1979).Another effect that parallels findings in PC is that EC is sensitive toUS revaluation. Baeyens, Eelen, Van den Bergh, and Crombez(1992) conducted a picture-picture study using photos of humanfaces. After an initial acquisition phase, the USs were revalued byaccompanying the liked picture with negative adjectives and the dis-liked picture with positive adjectives. At test, the ratings of the CSsshowed a shift in the similar direction as the shift in affective valueof the USs due to revaluation. This effect was still present at a one-month follow-up.

Discrepancies between EC and PC. Although EC appears toclosely resemble PC, there are some notable differences between ECand PC. For example, contiguity between the CS and US is importantin demonstrating PC. This is also certainly true for EC, but as op-

23

posed to PC, where the most learning is apparent when the CS di-rectly precedes the US (i.e., a delayed conditioning procedure), ECworks particularly well when the CS and the affective US are pre-sented in compound (i.e., a simultaneous conditioning procedure)(see Rozin et al., 1998). Further, in contrast with general findings inPC, EC can be demonstrated using a backward conditioning proce-dure, where the US precedes the presentation of the CS (Martin &Levey, 1978; Stuart et al., 1987). It should be noted though that thiseffect has not always been obtained (see Hammerl & Grabitz, 1993).

A more notable discrepancy with PC is the finding that statisticalcontingency does not play an important role in the transfer of affect.Baeyens, Hermans, and Eelen (1993) manipulated the degree of sta-tistical contingency between the CS and the US in a picture-picturestudy. Transfer of affect, indicative of EC, did not differ between agroup in which the CS was always paired with the US, a group inwhich the CS was paired on half of the trials with the US, and agroup in which the CS was paired with the US in only a third of thetrials. Closely related to this discrepancy is the repeated finding thatfor EC to take place, one does not need to be aware of the contin-gency between the CS and the US. Although there is still consider-able debate about the implicit nature of EC (see Field, 2000; Lovi-bond & Shanks, 2002), the general conclusion is that EC does not re-quire contingency awareness as opposed to PC (see De Houwer etal., 2001).

Another important discrepancy with PC is that EC appears to behighly resistant to an extinction procedure. Whereas in PC thelearned response can be extinguished by presenting the CS withoutthe US, such postacquisition nonreinforced exposure to the CS typi-cally does not affect the acquired affective value of the CS (e.g.,Baeyens, Crombez, Hendrickx, & Eelen, 1995; De Houwer et al.,2000). This does not necessarily imply that EC is different from PC.Pavlovian conditioned responses too cannot be extinguished in thesense that they are unlearned. When a Pavlovian CS is presented out-side the extinction treatment, conditioned responding typically re-covers (see Bouton, 1993). It has been argued that such a renewal ef-fect can also account for the apparent resistance of EC to extinction.Whereas the extinction of Pavlovian conditioned responding ismeasured during extinction treatment, the extinction of EC is usuallyassessed after an extinction procedure. The assessment of the affec-tive value of the CS at test after extinction may be regarded as anovel situation, or context, allowing for the renewal of EC (Lipp,Oughton, & LeLievre, 2003). Indeed, Lipp and colleagues (2003)demonstrated extinction of EC when assessing affective value of theCS during the extinction treatment. It should be noted though that theextinction of EC was less complete and far less rapid than the simul-taneous extinction of Pavlovian conditioned skin conductance re-sponding. Therefore, one can still argue that EC is exceptionally,though not completely, resistant to extinction.

Similar to the resistance to extinction, EC also appears to be resis-tant to modulation. In Pavlovian discrimination learning, responding

24

to a CS can come under modulatory control of a feature stimuluswhen the CS is reinforced in the presence of, or when preceded bythe feature stimulus (feature positive discrimination). Similarly, re-sponding to the CS can come under modulatory control when it is re-inforced only in the absence of the feature stimulus (feature negativediscrimination). The feature stimulus is said to set the occasion forthe CS to be reinforced (in feature positive discrimination training),or not reinforced (in the case of feature negative discrimination train-ing) (Holland, 1983; Rescorla, 1985). Baeyens and colleagues(Baeyens, Crombez, De Houwer, & Eelen, 1996; Baeyens,Hendrickx, Crombez, & Hermans, 1998) have repeatedly failed todemonstrate such occasion setting in EC.Conclusions. In many respects EC appears similar to PC. Nonethe-less, considering the specific characteristics of EC, particularly its re-sistance to extinction and the unimportance of contingency (aware-ness), EC is usually regarded as a form of associative stimulus-stimulus learning which qualitatively differs from PC (but see Davey,1994a). Since EC differs from PC, EC cannot be accounted for bywell-defined models of associative learning, such as the Rescorla-Wagner model (Rescorla & Wagner, 1972). Therefore, differentmodels of EC have been proposed to account for both the similaritiesand discrepancies with Pavlovian learning.

Models of ECThe conceptual categorization model of EC. Davey (1994b) has ar-

gued that EC does not reflect associative learning at all, but ratherconceptual categorization. According to Davey, a CS usually hassome features in common with the US it is paired with. Due to thispairing, the common features become more salient and thus the CS iscategorized as being more similar to the US. Field and Davey (1997)demonstrated how the pairing of stimuli promotes conceptual catego-rization. In this study, participants were presented with pictures offaces of Martians and Venusians. The researchers constructed thesefaces and the exemplar Martian and Venusian face differed on sixspecific features. In the experiment, ambiguous alien faces (sharingas many Martian as Venusian features) were paired with more proto-typical Martians or Venusians. At test, the ambiguous aliens wereevaluated as more Venusian-like when paired with a prototypicalMartian, whereas the ambiguous aliens that had been paired with theprototypical Venusian were rated as more Martian-like. Althoughthese results seem to be at odds with the general finding in EC thatthe evaluative shift is in the direction of the US, Field and Daveystate that these results nonetheless show how pairing of stimuli leadsto categorization and thus argue that such conceptual categorizationmay also play an important role in demonstrating EC.

Field and Davey (1999) point out that in several EC studies, CS-USpairs were constructed on the basis of perceptual similarity, hencepromoting conceptual categorization. Further, in a picture-picturestudy, they found no evidence for EC when the paired CSs and USswere perceived as dissimilar. Therefore, they argue that most demon-

25

strations of EC are not the result of associative learning, but the re-sult of an artefact due to the experimental paradigm (see also Shanks& Dickinson, 1990). De Houwer et al. (2001) however argue that inmany EC studies, CSs were assigned to USs on a random basis, orstimuli were counterbalanced, thus controlling for perceptual similar-ity effects. Baeyens, De Houwer, Vansteenwegen, and Eelen (1998)further argue that the conceptual categorization account of EC cannotexplain cross-modal EC. It is highly unlikely that the CS and the USare perceived as being similar in this situation. Furthermore, it cannotexplain why US revaluation should specifically affect the evaluationof the CS that has been paired with the revalued US (see De Houweret al., 2001).

The holistic representation model of EC. Martin and Levey (1994)do not discard EC as an artefact, but like Davey (1994b) they do notregard EC as the formation of an association. Martin and Levey de-scribe EC as the automatic formation of a holistic representation con-taining elements of both the CS and the US. When a neutral and anaffective stimulus are presented in compound, they are automaticallyfused or integrated into a single representation. According to theseauthors, this fusion process constitutes one of the most primitive andbasic forms of learning. In EC, the CS is thought to activate the ho-listic CS-US representation and thus the affective value of the US,accounting for the observed evaluative shift. It accounts for mostfindings in EC, but encounters some difficulty in explaining sensorypreconditioning as demonstrated by Hammerl and Grabitz (1996).Since the formation of a holistic representation depends on the co-occurrence of the CS and the US, a CS that has never been pairedwith the US should not be able to acquire affective value accordingto the holistic representation account.

The referential learning model of EC. Baeyens and colleagues(Baeyens, Eelen, Crombez, & Van den Bergh, 1992; Baeyens & DeHouwer, 1995) do describe EC in terms of the learning of an associa-tion, albeit different from Pavlovian learning. According to Baeyens,Eelen, Crombez, et al. (1992), human associative learning comprisestwo distinct learning systems: a signal learning system and a referen-tial learning system. The signal learning system controls PC. In PCone has to be aware of the contingency between the CS and the US tobe able to adequately anticipate the US when the CS is presented.However, in EC, controlled by the referential learning system, the CSmerely has to refer or activate the representation of the US for effec-tive transfer of affect. The referential learning model thus requires farless information processing resources and hence does not requiremuch conscious processing.

The referential learning model is very similar to the account ofMartin and Levey (1994), but as opposed to their account of EC,Baeyens and De Houwer (1995) do not regard EC as the basis ofmore complex stimulus learning such as PC. The referential learningmodel also explains most results of EC studies. However, Stevenson,Boakes, and Wilson (2000) point out that this model only allows af-fective value to transfer. Stevenson and colleagues (Stevenson, Pres-

26

cott, & Boakes, 1995; Stevenson, Boakes, & Prescott, 1998; Steven-son, Boakes, & Wilson, 2000) have repeatedly demonstrated that notonly affect, but other salient attributes of the US as well can transfer.For example, in one of their experiments, Stevenson et al. (2000)found that liking for a sucrose-paired odour was greater than likingfor a citric-paired odour. Next to the transfer of affect, the sucrose-paired odour was also rated as more sweet and the citric-paired odouras more sour at test. De Houwer et al. (2001) acknowledge that thereferential learning model cannot account for this result, unless it ispresumed that the referential learning system controls all but Pav-lovian anticipatory responding.

In summary, the present models of EC cannot account for all dem-onstrations of EC and are difficult to distinguish as they renderroughly the same set of predictions. De Houwer et al. (2001, p. 866)conclude that: “… these models tell us little about the specific proc-esses that underlie EC and thus do not permit the formulation of pre-cise hypotheses. It is therefore imperative that existing theories arereexamined and refined or that new models are proposed.”

Stimulus Generalization and Configural Learning as a Model of EC

In the first part of this article, we reviewed and summarized the dif-ferent findings concerning EC and discussed to what degree EC canbe conceptualized as a form of Pavlovian stimulus-stimulus learning(for a more comprehensive review concerning this topic we refer toDe Houwer et al., 2001). The present models of EC as describedabove all provide a relatively complete account of EC, but they can-not explain all findings and do not generate very precise predictions.Typically, these models either fail to account for the apparent impor-tance of perceptual similarity between stimuli in demonstrating EC,or fail to explain EC when paired stimuli are perceptually dissimilar.In reviewing the findings in EC studies we argue in the followingsections that EC is perhaps best understood in terms of stimulus gen-eralization and configural learning.

In 1987, Pearce devised a model of stimulus generalization, de-scribing the role of generalization and configural learning in the con-text of PC. Pearce presumes in accordance with Atkinson and Estes(1963) that one possesses a buffer containing representations of allstimuli present in the context to which one is exposed. According toPearce, when a stimulus is presented it activates a representation ofthat particular stimulus in the buffer. Through a process of spreadingactivation it can also activate representations of stimuli that are notpresented but share features in common with the presented stimulus.So when a CS X has acquired excitatory associative strength due topairings with a US (e.g., food), some conditioned responding will beelicited by a novel CS Y when this stimulus shares features in com-mon with X. This principle of stimulus generalization as formulatedby Pearce can also be applied to EC. According to the present modelthen, the hedonic shift observed in EC is due to generalization of af-fective value from the affective US to the CS. Similar to the gener-

27

alization of associative strength, we argue that the amount of general-ized affective value is determined by the perceptual similarity of theneutral stimulus with the affective US. Suppose that a stimulus X hasa certain affective value. The degree of generalized affect from X to astimulus Y then depends on the perceptual similarity of X and Y and,of course, the affective value of stimulus X. Equation 1 representsthis relationship and it directly corresponds with the equation of gen-eralized associative strength described by Pearce (1987).

.XYXY ASa ×= (1)

Parameter aY represents the total amount of generalized affectivevalue to stimulus Y. The value of aY depends on the similarity be-tween stimuli X and Y, represented by the parameter XSY, which canalso be described as the degree to which stimulus Y is able to directlyactivate a representation of X. Affective value of stimulus X is repre-sented by the parameter AX and can take on a value anywhere be-tween –1 (negative valence) and 1 (positive valence).

The S parameter in Equation 1 requires some more specification. Inaccordance with Pearce (1987, 2002) we argue that the similarity (S)between two stimuli depends on the number of common features andthe total number of activated representations by each stimulus.Pearce (1987, 1994) formulates similarity as displayed in Equation 2.

.Y

C

X

CYX n

nnnS ×= (2)

Equation 2 describes how S depends on the ratio of the number ofcommon features nC (not only including the perceptual features of thestimuli, but also the context in which these stimuli are presented) andthe total number of activated representations by stimulus X, as wellas the ratio of the number of common features and the total numberof activated representations by stimulus Y. If one assumes that stimu-lus X and Y are equally intense and thus activate the same number ofcommon representations, Equation 2 can be rewritten as Equation 3.

)(

2

YX

CYX nn

nS×

= (3)

Although formulated somewhat differently than by Davey (1994b),we also argue that the perceptual similarity between stimuli plays akey role in demonstrating EC. Using a picture-picture paradigm,Field and Davey (1999) demonstrated the importance of perceptualsimilarity in EC. Pictures of human faces were evaluated after whichneutral pictures (CSs) were paired with either liked or disliked pic-tures (the USs). At test, participants had to rerate the pictures. Shiftsin affective value were found, but only for the CSs that were catego-rized as being similar to the US they had been paired with. Moreover,

28

such transfer of affect was also found in a control group that had re-ceived explicitly unpaired presentations of the CSs and the USs andin another control group of participants who simply had to rate thepictures twice. Note that the participants in these control groups werenot informed beforehand that pictures could or could not be paired.

Although perceptual similarity certainly plays an important role inEC, De Houwer et al. (2001) emphasize that perceptual similarity isnot a prerequisite for demonstrating EC. The stimulus generalizationmodel of EC as presented thus far, also cannot provide an account ofEC when the CS and the US are perceptually dissimilar. However,the generalization of affective value can be promoted through theformation of a configural representation of the CS and the affectiveUS. The potentially important role of configural representations inhuman associative learning has already been noted by several re-searchers (e.g., Shanks, Charles, Darby, & Azmi, 1998; Shanks,Darby, & Charles, 1998; Stevenson et al., 2000).

According to Pearce (1987), a configural representation is formedwhen two or more stimuli are presented in compound. Pearce (2002)has described how such configural representations may play a role inanimal flavour-flavour learning, a phenomenon that bears close re-semblance to the EC paradigm. In animal flavour-flavour learning,animals receive pairings of artificial flavours paired with an inher-ently preferred flavour (e.g., a sweet flavour). This leads to the ac-quisition of a conditioned flavour preference for the previously neu-tral flavour. As in EC, conditioned flavour preferences are acquiredvery rapidly and are highly resistant to an extinction procedure (seeCapaldi, 1996). Pearce states that whenever a neutral flavour ispaired with the preferred flavour, the co-activation of the representa-tion of the neutral and preferred flavour leads to the formation of aconfigural representation, comprising elements of both flavours.When the neutral flavour is presented at test, a representation of thepreferred flavour is activated indirectly through the activation of theconfigural representation. The degree to which the initially neutralflavour is able to activate the configural representation rapidly in-creases, requiring but a few trials.

XSY

YSXY

XYSX

X

Y

XY aY = (XSY + (YSXY × XYSX))× AX

29

Figure 1. A schematic representation of how an initially neutralstimulus Y acquires affective value from an affective stimulus X.Due to pairing of X and Y a configural XY representation is formed.When Y is presented, affective value of X (AX) will generalize to Y(aY) to the extent that Y is capable of directly activating a representa-tion of X (XSY) and activating the representation of X via the con-figural XY representation (YSXY × XYSX).

Figure 1 displays how this notion of configural learning can also beapplied to EC. As described by the figure, when an affective stimulusX is paired with a neutral stimulus Y, the co-activation of the repre-sentation of X and Y leads to the formation of an XY configural rep-resentation. When Y is presented at test, it activates the representa-tion of X directly to the extent XSY, and indirectly through the activa-tion of XY to the extent YSXY × XYSX. As such, the stimulus generali-zation model does not preclude the possibility of cross-modal EC, incontrast to Davey’s model (1994b). However, it does predict thatcross-modal EC is more difficult to demonstrate. Further, it predictsthat effective EC requires but a few CS-US pairings.

Summarized, affective value can generalize both directly and indi-rectly from multiple affective sources. Therefore, Equation 1 can beextended to the more general Equation 4, which again directly corre-sponds with Pearce’s 1987 notion of generalization of associativestrength in the context of PC. Equation 4 provides an equation for de-termining the degree of generalized affective value to stimulus Yfrom n stimuli similar to Y having some affective value (positive, ornegative).

∑ ×= .nYnY ASa (4)

The present stimulus generalization model of EC explains the irrele-vance of contingency in EC by stating that EC partly depends on theformation of a configural representation. This latter process of con-figuring is thought to occur automatically relying solely on the co-activation of the representations of the presented stimuli. Pearce’s1994 connectionist model of configural learning describes the forma-tion of a configural representation as an automatic and immediatenon-iterative process, as opposed to the gradual formation of a Pav-lovian association between stimuli. Differing levels of contingencyshould thus exert no effect on the demonstration of EC and one neednot be aware of the specific contingency between the CS and the US. The model also predicts that simultaneous presentations of the CSand the US should be most effective in demonstrating EC. AlthoughRozin et al. (1998) note that this seems to be the case, EC has beenfrequently demonstrated using forward pairings of the CS and theUS. In its present form the stimulus generalization model of EC does

30

not permit the formation of a configural representation when the CSand the US are presented sequentially. As Pearce (2002) notesthough, the model can accommodate the formation of a configuralrepresentation when stimuli are presented in a serial compound if it isassumed that activation of the representation of a stimulus is not ter-minated instantly with the offset of the presented stimulus, butgradually decays.

In summary, the present model predicts that perceptual similarityplays a key role in demonstrating EC. It also predicts that contiguity,but not contingency, is important in demonstrating EC. Apart formthese initial predictions, the stimulus generalization and configurallearning model of EC can account for most typical EC findings.

Application to extinction. As the model describes EC in terms ofstimulus generalization and configural learning, the model predictsthat the acquired affective value should be extremely resistant to ex-tinction. Once a configural representation is formed, nonreinforcedexposure to the CS should not affect its ability to activate a represen-tation of the configural CS-US representation, and hence generaliza-tion of affective value should be relatively unaffected by an extinc-tion procedure.

Application to modulation. The model predicts that EC should re-sist modulation as Baeyens and colleagues (Baeyens et al., 1996;Baeyens et al., 1998) demonstrated. In one of their experiments (seeBaeyens et al., 1996) the participants tasted a flavour (A) having acertain colour (X) paired with Tween (a taste which was generallyrated as highly aversive by the participants). When the flavour waspresented without the specific colour, the flavour was not paired withTween. Participants thus received two trial types: a type in whichflavour A was not paired with Tween (A0 trials) and a trial type inwhich the flavour-colour compound XA was paired with Tween(XA+ trials). Colour X did not come to modulate evaluative respond-ing at all. At test, evaluative shifts were observed for the flavour Awhether or not presented in compound with the colour. The modelexplains this pattern of findings by suggesting that during discrimina-tion training an XA-USTween configural representation was formed.Due to X being of a different sensory modality than both A and theUS, the similarity between X and XA-USTween would be muchsmaller than the similarity between A and XA-USTween. Generaliza-tion of affective value from the US would thus be limited primarilyto stimulus A and the XA compound as was demonstrated byBaeyens and colleagues (1996).

Application to counterconditioning. Baeyens and colleagues (1989)found that EC is sensitive to a counterconditioning procedure. Al-though at face value this finding appears to imply that EC reflectssome form of associative learning, the present nonassociative modelof EC can accommodate this pattern of results. According to themodel, in the initial acquisition phase a configural representation isformed containing both elements of the CS and the affective US1.When in the following phase the CS is now paired with another af-fective stimulus (US2) that has an affective value directly opposite to

31

US1, presentation of the CS at test will activate a representation ofboth USs. Assuming that the CS activates the USs to a similar extent,the net generalized affective value to the CS will approach zero.

Application to US revaluation. Baeyens, Eelen, Van den Bergh, etal. (1992) showed that EC is sensitive to US revaluation. The presentmodel can accommodate the observed US revaluation effect withoutreferring to the formation of an association. The model predicts thatwhen an affective US is revalued, the CS that has previously beenpaired with this US will show an evaluative shift in the same direc-tion as the US, because at test it activates the representation of therevalued US through the activation of the previously acquired con-figural CS-US representation.

Application to sensory preconditioning. As described above, thefinding of evaluative sensory preconditioning (Hammerl & Grabitz,1996) posed serious implications for the holistic representation ac-count of EC proposed by Martin and Levey (1994). Although thepresent model of EC is very similar to the notion of the formation ofa holistic representation put forth by Martin and Levey, it does notpreclude sensory preconditioning. In the first phase of the paradigm,it is predicted that a configural CS1-CS2 representation is formed. Inthe subsequent conditioning phase, a CS2-US configural representa-tion will be formed. When CS1 is presented at test it will be able toactivate the representation of the US directly through its perceptualsimilarity with the US and indirectly through the activation of theconfigural representations.

Application to stimulus preexposure. Another effect in EC that ap-pears to demonstrate the associative nature of EC is the detrimentaleffect of stimulus preexposure on EC (e.g., Stuart et al., 1987; Ham-merl et al., 1997). The present model cannot explain the detrimentaleffects of CS (or US) preexposure on EC, unless it is assumed thatsuch preexposure leads to perceptual habituation to the preexposedstimulus, hence limiting the formation of a configural representationduring the subsequent conditioning phase. Hammerl et al. have madea somewhat similar proposal in explaining the effects of US exposurein EC. In concordance with Randich and Lolordo (1979), Hammerland colleagues state that repeated US exposure leads to habituation tothe US and as a consequence the US will loose some of its affectivevalue. As such, both preexposure and postexposure of the US shouldattenuate EC.

Although the present model of EC encounters some difficulty inexplaining stimulus preexposure effects on EC, it should be notedthat these effects have not been studied extensively as of yet. More-over, the proposed detrimental effect of CS preexposure on EC hasnot been found to be a reliable effect (see De Houwer et al., 2000;Stevenson et al., 2000).

Tests of the Model As described above, the stimulus generalization and configurallearning model of EC accommodates most EC findings. However,the question arises how to test the present model. The present model

32

of EC states that any evaluative shift results from stimulus generali-zation and not as a result from predicitive learning, that is PC. There-fore, it should be possible to dissociate these accounts of evaluativeconditioning.

Discrimination learning. One of the primary assumptions of thepresent stimulus generalization model is that similarity betweenstimuli is an important factor in demonstrating EC. In EC a CS willshow an evaluative shift in the direction of the affective value of theactivated representation of the affective US it has been paired with. Ifa CS is thus paired with both a positive and negative stimulus, theevaluative shift at test will be determined by the degree in which therepresentations of these different USs are activated. When both USsare activated to the same extent and have directly opposing affectivevalues, no EC will be apparent at test.

However, consider the following discrimination. A neutral stimulusA is paired with a negative US, another stimulus B is paired with apositive US. Next to these paired presentations, A is also presented incompound with another CS C and paired with the positive US, and Bis also presented in compound with C and paired with a negative US.This then leads to an A – US1

-/B – US2+/AC – US2

+/BC – US1- dis-

crimination. How would such a discrimination affect the transfer ofaffect to A, B, and C separately?

The referential model states that an evaluative association dependssolely on the co-occurrence of the CS and the US. As each CS in thiscase co-occurs with both a positive and negative US, no significantEC is expected to be observed according to this model. The concep-tual categorization model renders the same prediction when one as-sumes that all individual stimuli are equally similar to one anotherand the positive and negative USs have directly opposing affectivevalues. The holistic representation model does not specify when orhow a particular holistic representation will be retrieved and thusdoes not render a specific prediction in this case. In contrast to theseexisting models of EC, the stimulus generalization model predictsthat one can ‘solve’ the discrimination. It predicts that A shows anegative shift, B a positive shift and no substantial evaluative shiftfor stimulus C. Presentation of A is more likely to lead to the activa-tion of the representation of the negative US, as A is more similar tothe configural A – US1

- stimulus than the configural AC – US2+ rep-

resentation. In contrast, B will more strongly activate a representa-tion of the positive US, as stimulus B is more similar to the con-figural B – US2

+ stimulus than the BC – US1- representation. Presen-

tation of C will activate the representation of both the positive andnegative US to a similar extent, thus limiting EC. Whether one trulycan ‘solve’ such a discrimination remains to be investigated.

Compound conditioning. Another prediction of the present modelconcerns compound conditioning. In PC, when two CSs are pre-sented in compound and paired with a US, this compound stimuluswill be just as effective at eliciting conditioned responding as anyother single CS paired with this US. The present model, however,predicts that when a compound comprising two neutral CSs (e.g., A

33

and B) is paired with an affective US Y, the transfer of affectivevalue to this compound will be less than the transfer of affectivevalue from Y to a CS comprising a single neutral stimulus X. Whenthe AB compound is presented at test, this will activate the AB-Yconfigural representation that in turn activates the A, B and Y repre-sentations. As the similarity between the AB-Y representation and Yis smaller than the similarity between the X-Y representation and Y,Y will be less activated by presentation of the AB compound ascompared to X and thus, the transfer of affective value from Y to theAB compound will be less than the transfer from Y to X. Addition-ally, the AB compound also activates representations of its neutralelements A and B separately, and this further limits the transfer of af-fective value. None of the other models of EC make such a predic-tion, but again, whether compound conditioning is indeed different inEC in comparison with PC awaits further research.

Learned irrelevance. In PC, when a CS is explicitly unpaired withthe US, this typically retards the formation of an association whenthe CS is later made predictive of the US. This effect has beentermed learned irrelevance and can be explained by arguing that dur-ing the uncorrelated presentation of the CS and the US, the context inwhich the CS and US are both presented acquires associativestrength, hence blocking the later acquisition of a CS-US associationwhen the CS is made predictive of the US (see Balsam & Tomie,1985). If EC can be explained in terms of PC, then one would expectto be able to demonstrate such an effect within an EC paradigm.However, the present model predicts that – given that the CS and USare to some degree perceptually similar – even uncorrelated pairingsof the CS and US may already lead to a shift in the subsequentevaluation of the CS. Therefore, prior uncorrelated presentations ofthe CS and an affective US promote rather than attenuate subsequentevaluative learning when the CS and affective US are presented con-tingently.

Context specificity. Pavlovian conditioned responding can be con-text specific. For example, extinguished conditioned responding canbe renewed when the CS is presented in a context different from thecontext in which the extinction treatment took place. This impliesthat extinction reflects inhibitory conditioning and that this secondlearned inhibitory meaning of the CS is context specific. Similar con-text specificity applies to Pavlovian excitatory conditioned respond-ing when inhibition to the CS has been learned prior to excitatoryconditioning training (Nelson, 2002). Generally though, initially ac-quired Pavlovian conditioned responding generalizes easily acrossdifferent contexts (Bouton, 1993). This, however, should not be thecase with EC. A CS and an affective US are always paired within aspecific context. Therefore, according to the present model at least,features of this context will be incorporated in the configural CS-USrepresentation. As such, presenting the CS in another context shouldalways lead to generalization decrement and hence a loss of an ini-tially acquired affective shift.

34

It’s important that you aim to interest, inform and persuade your readership (i.e., yoursupervisors probably). Try to express yourself as clearly and rationally as possible andbase your arguments solely on empirically relevant observations (Fernandez-Rios &Buela-Casal, 2009). Avoid emotional appeals and reasoning; avoid speculations thatgo much beyond the available research data, and you should take pain to avoid com-mitting logical fallacies. The website www.logicalfallacies.info provides a very in-formative and useful taxonomy of logical fallacies and reasoning errors. Based onyour research question you probably have a fair expectation of what answer you willfind in the research literature. Don’t avoid reading or reporting the papers that supporta conclusion opposite to your expectation!

The end of your review paper concerns discussion and conclusions. Fernandez-Rios and Buela-Casal (2009, p. 333) give some excellent advice on organizing the endof your paper:

“Depending on the type of content and organisational structure of the article, conclu-sions can be drawn in one or two paragraphs. Conclusions should not be a simplerepetition of results. They should clearly state the theoretical and practical implica-tions for the future. There should also be a relationship between results, discussionand conclusions… .”

Here’s the example for the evaluative conditioning paper:

Concluding Comments

The primary aim of the present article has been the development of anew model of EC. The model accommodates most EC findings andgenerates specific hypotheses concerning EC effects. It can be arguedthat previous models of EC either fail to explain the importance ofperceptual similarity in demonstrating EC (e.g., Baeyens, Eelen,Crombez, et al., 1992; Martin & Levey, 1994), or fail to explain thepossibility of the transfer of affect when stimuli are perceptually dis-similar (Davey, 1994b). By describing EC in terms of generalizationof affective value and emphasizing the role of configural learning insuch generalization, the importance of perceptual similarity is incor-porated in the present model while not precluding the possibility ofEC when stimuli are perceptually dissimilar. As such, it can be ar-gued that the present model provides a more comprehensive accountof EC than previous models have.

One may argue that perceptual similarity lies in the eye of the be-holder. Judging similarity between two stimuli may differ betweenspecies and even between individuals of the same species. Describingperceptual similarity in terms of the number of common elements be-tween stimuli thus appears to be an oversimplification. Nonetheless,

35

it allows for the simple manipulation of perceptual similarity. In-creasing the number of common elements between two stimulishould make these stimuli more alike for each individual member ofeach type of species. Therefore, within an individual subject increas-ing the number of common elements between the CS and the affec-tive US should always lead to more effective evaluative learning.As noted above, the present stimulus generalization and configurallearning model of EC is an adaptation of the 1987 Pearce model ofstimulus generalization in the context of Pavlovian learning. Sincethis model can be so easily modified to account for both the generali-zation of associative strength and affective value, there is no reasonto assume that stimulus generalization should be restricted to thegeneralization of these two potential stimulus attributes. So althoughthe generalization model is presented here as a model of EC, it canalso explain the simultaneous transfer of other salient stimulus attrib-utes. Although the stimulus generalization and configural learning ac-count of EC provides a more complete and specific account of EC, asany other model it requires rigorous testing. What it has in commonwith other models of EC is that it describes EC as being distinct fromPC. This means that in specific cases it should be possible to dissoci-ate EC from PC. In this respect, the present model renders more spe-cific predictions than other previous models of EC. We agree withDe Houwer et al. (2001) that future EC studies should be explicitlyaimed at testing the boundary conditions of EC. Such studies shouldprovide results with which one could evaluate the merits of the pre-sent model.

Overall article structure

The review paper introduction is preceded by an abstract, just as with a research paper(describing the paper in a single paragraph no longer than 150 words). Write the ab-stract after you’ve completed the rest of the article as it is based on that content. Donot insert any references or subheadings in your abstract, but you may add keywordsto the abstract (though consult with your supervisor). For example:

Abstract

Evaluative conditioning refers to the transfer of affective valueto an initially neutral stimulus by pairing the neutral stimuluswith an affective stimulus (positive, or negative). In the first partof this article the associative nature of evaluative conditioning isdiscussed and existing models of evaluative learning are de-scribed. It is argued that these models of evaluative conditioningdo not provide a complete account of evaluative conditioningand generate non-specific predictions. The second part of the ar-ticle concerns the development of a more specific model ofevaluative conditioning, describing such conditioning in terms of

36

Pearce’s 1987 model of stimulus generalization and configurallearning. By viewing the transfer of affect as a generalizationand configural learning process, this model can account for mostdemonstrations of evaluative conditioning and generates moreprecise and unambiguous predictions.

Key words: affect; configural learning; evaluative conditioning;generalization; Pavlovian conditioning

The title page precedes the abstract. It contains a brief title that is attractive and in-formative. On your student review paper you also need to provide an overall wordcount (how many words does your paper count? – make Word do the counting foryou) your full name(s) and student ID, the name of your supervisor/mentor, and thatyou’re a student at Maastricht University, Faculty of Psychology & Neuroscience.

Here then is the general structure of a review paper:

37

INTRO (page 3)(What’s the problem? What question

will be answered? What are theaims?)

DISCUSSION &CONCLUSIONS

(Discuss your findings. Draw conclusionsregarding your original question and identify

new questions/problems and avenues forresearch)

MIDDLESECTIONS(What did you do toanswer the researchquestion? And what

did you find?)

TITLE (page 1)

ABSTRACT (page 2)

REFERENCES (new page)

38

Style of writing

As a researcher you have an important message to convey and that message concernsyour brilliant research results. If you want to convince your peers of your brillianceyou need to be able to tell your ‘story’ as clearly and as grippingly as possible.

Write clear and simple sentences. Formulate unambiguous sentences and do thisas precisely as possible so that it is crystal clear what you mean. Balance your argu-ments and be honest. You may try and convince your reader, but not by bluffing or bybiased presentation of the research literature. Use a professional tone. Be polite anddo not litter your text with silly jokes. You may criticize other people’s research, butnever the individuals who did that work. In his essay Politics and the English Lan-guage, George Orwell (author of Animal Farm and 1984) provides some very goodtips on writing clearly:

• Never use a metaphor, simile or other figure of speech you are used to seeingin print

Indeed, not all of your readers are native speakers of English. Using metaphors,proverbs and sayings in your writing will muddle its clarity.

• Never use a long word where a short one will doFor example: Self-administation of psychotropic substances = drug use

• If it is possible to cut a word out, always cut it out• Never use the passive where you can use the active tense

It was investigated whether...(passive)We investigated whether... (active)This is good advice, but it clashes with the idea of employing a professional and ob-jective tone of voice. Some scholars believe that a paper becomes or appears too sub-jective when written completely in the active tense using personal pronouns such as‘I’ or ‘We’. You better check with your supervisor how s/he feels about writing sen-tences in active tense; some supervisors have no problem with that, whereas othershave.

• Never use a foreign phrase, a scientific word or a jargon word if you can thinkof an everyday English equivalent

This may appear as a bit of stupid advice as most of the writing you’ll be doing dur-ing your educational training will concern science writing, but there is a lot of jargon(‘psycho-babble’) in psychology and even your supervisor may not be familiar withsome of the professional jargon that’s being used within a specific subject area. So atthe very least, ALWAYS define concepts and terms and do not hesitate to use exam-ples in making clear what you exactly mean by the term.

Abbreviations in textDo NOT use abbreviations, apart from:

39

1. Statistical terms:M meanSD standard deviationDf degrees of freedom

2. Standard measurement units:cm centimeterg gramml milliliterms millisecond

3. Standard terms:RT reaction timeSTM short-term memoryANOVA analysis of variance

TransitionsWhen you’re writing a paper it helps to remind yourself that the reader of your paperwill be able to better understand your paper if s/he can discern some logical flow orline of reasoning. To accomplish this, it helps if you use transition phrases betweensections and paragraphs. These phrases can make use of certain words that signal sucha transition expressing a conclusion (e.g., adverbs(Un)fortunately,…Similarly,…Certainly,…Clearly,… ), or contrast (between sen-tences: By contrast,…On the other hand,…However,..), or a series (First,… Second,…Third,…; In addition…Additionally…Then…; Last,… Finally,… Also,…; Fur-ther,…Moreover,… Another… ) or implications (Therefore,… ; It follows that… ; Inconclusion,… ; If… then… ; … thus… (use within sentences); … hence… (use withinsentences).

Common mistakes of ‘a’ and ‘e’If you’re writing an assignment in English, the APA Publication Manual warns ofsome very common mistakes writing ‘a’ when one should have written an ‘e’ and viceversa.Affect or Effect? Affect as a verb means ‘to influence’. Affect as a noun refers to emo-tion or feeling. Effect refers to an outcome or result of something.Then or Than? ‘Than’ is always used in comparisons (William James is stronger thanWilhelm Wundt). ‘Then’ should be used as a time reference (James then kickedWundt in his teeth).

Singular - pluralDatum – Data; Stimulus – Stimuli; Analysis – Analyses; Phenomenon – Phenomena;Hypothesis – Hypotheses

40

Formatting

Your paper should be carefully prepared according to APA style. This may not appearvery important, but there are supervisors among the teaching staff who are very strictabout the application of APA style guidelines. Follow these guidelines as closely aspossible. But as your paper is not intended for publication in a science journal, as itserves as an educational assignment, there are some specific formatting guidelines foryou to adhere to that do not necessarily correspond with the overall APA guidelines.

1. For your paper you need to use a clearly legible font.2. Use MS Word for writing and preparing your paper in and use the standard

Times New Roman 12 pts font.3. You’re allowed to use 16 pts font size for the title.4. Throughout your paper use 1.5 line spacing.5. Number your pages throughout your paper. The title page is page 1 unless

stated otherwise.6. The paper margins (upper, lower, right, and left) should be kept at 2.5 cm.

41

Presenting research in PowerPoint

Communication of research is not confined to writing an article. Another way tocommunicate your research is through giving a talk at some big science conference.Usually such a presentation concerns a MS PowerPoint slide show. So how shouldyou prepare such a slide show presentation? I once attended a presentation at a con-ference in Spain and the presenting scientist had his whole presentation written out onpaper, which he read out loud. He might as well not have been in the room, becausethe entire presentation (word for word) was presented on a series of PowerPointslides. It was like watching a very poor karaoke performance of an extremely long-winded song. But it wasn’t karaoke; it was one of the worst possible research presen-tations. So here are some tips and tricks for you to do a much better job.

Start with preliminary information to make it clear and explicit what you are talkingabout. The most important part of this preliminary information concerns your researchquestion. What question did you try to answer and why is that an important or inter-esting question? Don’t waste too much time on preliminaries or else you won’t haveenough time for the good stuff. The good stuff is the answer to your research ques-tion. Put this in the middle of the talk. Save the more boring stuff (study limitationsand considerations for future work) for the end. Keep your summary and conclusionsbrief. Only emphasize the most important concepts, the things you want your audi-ence to remember.

MS PowerPoint is used by most researchers to prepare their talk in. I presume thatyou are already somewhat familiar with working with PowerPoint. Still, that doesn’tmean that you already know how to organize your slides efficiently.

1. People cannot absorb too many facts from what is usually a brief talk (about15 min). Therefore, you need to focus on the key points of your research pa-per. Your slides are intended to help the audience grasp those points.

2. Keep the slides simple and easy to read. In case of a longer talk (more than 20min); it may be useful to show an outline of your talk on a slide somewhere atthe beginning of your presentation.

3. Use large fonts (24 pts or even larger) and don’t try to fit more than 6 lines oftext on each slide.

4. Show something on each slide and don’t waste space with putting large logosof Maastricht University or whatever on them; these logos distract from thecontent plus you’re not presenting some marketing message.

5. Sans-serif fonts such as Arial are more readable, especially for audience mem-bers sitting in the back of the room.

42

6. You can use different text colours but always keep in mind to keep the slidetext readable and that your audience doesn’t care about the visual aesthetics ofyour presentation.

7. PowerPoint provides you with the option to use special effects with slide tran-sitions or the appearance of text lines on the slides. Apply these effects spar-ingly, if at all. They distract and generally use up too much time.

8. Don’t put boring stuff on your slides, such as statistical values, or literaturereferences.

When presenting, you are talking to an audience of listeners. If you want to effec-tively communicate your message you must speak clearly. This will make you feellike an idiot if not a little bit self conscious. Don’t worry, that’s normal.

1. Don’t speak too fast and don’t speak in a soft murmuring voice.2. Don’t fiddle with your hands or hold your hands in your pockets or behind

your back. Use hand gestures to emphasise your speech, your arguments andimplications.

3. Don’t stare at the ceiling during your talk, don’t stare at the floor, and don’tstare at your piece of paper outlining your presentation. Engage your audienceby regularly making eye contact with them.

4. Practice your presentation as many times as you can beforehand and do thisout loud. Practice makes perfect.

43

OVERVIEW OF

WRITING/COMMUNICATION

ASSIGNMENTS IN THE FPNPSYCHOLOGY BACHELOR

CURRICULUM

44

Skills II

The Skills II writing assignment

You and your fellow group members have conducted a small observational experi-ment. You have conducted the necessary statistical analyses in a special SPSS practi-cal session and you have been taught to find research articles in the enormous elec-tronic databases subscribed to by the university library here in Randwijck. Ergo, withthe above manual you have acquired all the necessary tools to write your paper con-cerning your research! Good luck!

The paper should have the structure outlined above in the manual and have a totallength of at least 7 pages and a maximum length of 10 pages. The paper should in-clude at least 3 relevant references (preferably to journal articles; be smart and searchfor literature review articles). Try to find your own way in the library, but do not hesi-tate to ask your supervisor/mentor for any help if you so happen to get lost. You canwrite the paper in either DUTCH or ENGLISH.

The Skills II presentation assignment

You and your fellow group members have conducted a small observational experi-ment. You’ve written a paper on that research for the Skills II writing assignment.Now you need to present your research in a 12 min presentation (note: 1 slide = 1min; a 12 min talk will give you enough time for presenting 12 slides). You do thispresentation with your group members.

The outline of your presentation should closely follow the research paper: Introduc-tion; Methods, Results, Discussion. Have one group member present the introduction,another one present the methods and/or results and another one the discussion. Youmay present your paper in either DUTCH or ENGLISH but note that the entire pres-entation needs to be in one language only.

45

Skills III

Written assignment 1 (PSY1134)

Select and describe one of the following concepts without directly mentioning or re-ferring to the concept in 100 words (min 80 – max 120 words). This descriptionshould make the concept crystal clear. The concepts themselves are borrowed fromthe course Development and the course Perception, so you should be familiar withthese concepts and it shouldn’t cost you too much trouble to write an accurate descrip-tion.

Why is this brief exercise important? By now you probably know that psy-chology contains a lot of jargon. It almost has its own language. It’s a language thatmost lay people do not understand. Writing is about communicating your ideas andthoughts to a readership. Even when the reader is a psychologist, you can’t presumethat s/he will understand every psychological concept or jargon word you use. It’sbest to keep the jargon limited to an absolute minimum. But sometimes you cannotsimply ignore concepts, key concepts in your writing. For example, if you write apaper on ‘cognitive dissonance’, it makes little sense not to mention cognitive disso-nance. However, what you then should do is clarify to the reader what you exactlymean by cognitive dissonance. This is the exercise.

Please beware! Don’t make this a copy-(from Wikipedia, from articles, fromother students, or from course books) cut-and-paste assignment. We check every writ-ten assignment for such fraud. If you’ve simply copied text, you haven’t learned any-thing and you’ll fail on this assignment.

Concepts• Cognitive variability• Adolescence• Folk psychology• Attachment• Theory of Mind• Fechner’s Law• Colour constancy• Fourier analysis• Auditory scene• Recognition by Components Theory

46

Example

Written assignment 1 (PSY1134) CONCEPTStudent: Remco HavermansStudent ID: 13101974Concept: cognitive dissonanceWord count: 99

In the 1950s, Leon Festinger proposed that we have a built in mechanism that makesus feel uncomfortable whenever we experience an inconsistency between our attitudesbeliefs, feelings, and thoughts. One may really enjoy smoking cigarettes, while at thesame time knowing that smoking these cigarettes is terrible for one’s health. This in-consistency will induce discomfort and one becomes motivated to reduce and avoidthat distress by convincing oneself that smoking only a few cigarettes per day is notall that harmful and by ignoring people who tend to tell every smoker that smoking isa filthy, unhealthy habit.

Gray, P. (1999). Psychology. New York: Worth Publishers.

Written assignment 2 (PSY1135)

The procedure section in any empirical paper is a description of what a participant hadto do or was subjected to in the experiment. It is an important section of the researchpaper as it gives the reader insight into how the experiment was conducted and how itwas designed. It thus provides the reader with the necessary information to criticallyevaluate the merits of the study and how to replicate it.

As a student, you are often approached to take part in an experiment. In fact,you earn credits by taking part in experiments and participating in research is thus astudy requisite. This assignment is an exercise in writing a procedure section. As youhaven’t conducted any experimental research by yourself your job in this exercise isto describe a procedure from your own experience as a participant. If you haven’tbeen a research participant already, then be one now. Describe your experience inchronological order in a minimum of 300 words and a maximum of 350 words. Payattention to details. Were you contacted by the experimenter beforehand? Was theexperimenter male or female? Where did you meet? Where were you tested? Whatwas the room or lab like? What did you have to do first? Did you receive any specificinstructions? Were you paid for your participation? Were you debriefed?

Don’t forget to put yourname and ID on your as-

signment.

Use Times NewRoman 12 pts and1.5 line distance

Provide the refer-ences you used.Use at least 1reference! UseAPA style forreferencing!

47

The procedure is not a reflection of how you felt about participating in theresearch. Do not describe your personal thoughts and feelings you experienced whenparticipating. Stick to the bare facts and use a professional objective tone. I under-stand that this may sound a bit abstract, so I’ve added an example of what I expectfrom you below.

Please beware! Don’t make this a copy-(from Wikipedia, from articles, fromother students, or from course books) cut-and-paste assignment. We check every writ-ten assignment for such fraud. If you’ve simply copied text, you haven’t learned any-thing and you’ll fail on this assignment.

Example

Written assignment 2 (PSY1135) PROCEDUREStudent: Remco HavermansStudent ID: 13101974Word count: 347

I recently took part in an experiment. I was welcomed by a female experimenter andshe escorted me to a separate room. She let me read written information about theexperiment. I was asked to sign a consent form if I still wished to participate. I signedthe form and was then given a set of questionnaires to fill out. Most of the questionshad to do with how much I worried about my weight and whether I tend to eat when-ever I feel sad.

I was tested alone. I sat in a quiet laboratory room measuring approximately10 m2. Before testing, the experimenter measured my weight and height. She ex-plained to me that she wanted to determine my body mass index, but she did not tellme why. I first received a test to assess how well I can smell. I was blindfolded andthe experimenter held different smells under my nose. I had to say whether I couldsmell anything, and I was asked to identify different smells such as pineapple orleather. Next, the blindfold was removed and I had to taste four different foods:Speculoos, cheese spread, a cinnamon cookie, and a cheese cookie. The Speculoosand cheese spread were served on a plastic tablespoon, and I was instructed to justlick it off the spoon.

For each food I had to rate pleasantness of flavour on a scale ranging between0 ‘‘highly unpleasant’’ to 100 ‘‘highly pleasant.’’ Next, I received 5 cinnamon cook-ies. The experimenter instructed me to take a cookie from a serving tray and then lookat it for 10 seconds, sniff at it for 10 seconds, to chew and finally to swallow it. WhenI finished all 5 cookies I again received the four foods to eat and evaluate. Followingthese final ratings, I was thanked and I received a 10 euro voucher (VVV bon/ IRIScheque). The experimenter told me afterwards that the purpose of the experiment was

48

to examine to what degree olfactory acuity determines the development of satiation tothe cookies I had to eat.

Written assignment 3 (PSY1136)

A very important part of writing a psychological science paper constitutes reading.Never make the mistake that the brilliant insight you had when lying in your bedthinking about science all night long (yes there are students who do that), has neveroccurred to anyone else before. It has, and you need to acknowledge that in your re-search paper. It goes without saying that when you’re writing a literature review youalso need to read and understand the relevant body of research literature. Especially inthe latter type of paper you need to be able to succinctly describe key studies in such afashion that it shows (the reader) that you have read and now correctly understand theresearch.

In this assignment you’ll have the opportunity to systematically read andsummarize a classic psychology research paper, classic in the sense that these are oldbut still valid and have advanced the field in important ways. The specific field I’mreferring to is social psychology. So you need to summarize a social psychology clas-sic, that is, one of three classic papers in 350-500 words. The choice for specificallysocial psychology research is twofold: 1) social psychology has a long history ofclever experimental work and 2) social psychology truly is “everybody’s business”(Myers, 2002).

Please beware! Don’t make this a copy-(from Wikipedia, from articles, fromother students, or from course books) cut-and-paste assignment. We check every writ-ten assignment for such fraud. If you’ve simply copied text, you haven’t learned any-thing and you’ll fail on this assignment.

The summary you’ll make will be based on a structured reading of the article. How toread an empirical article?1 Just use the reading tool below.

General aspects of the paperQuestion Place to lookWhat is the aim of the paper?

What question is addressed?

abstract, end of the introduction, finalparagraph discussion;first paragraph of discussion

What experimental paradigm/task isused? Has it been used before, and whatis the typical interpretation? What under-lying processes are being 'tapped' by the

Introduction and method

1 Fren Smulders originally devised this reading tool.

49

paradigm?What is the general hypothesis? end of the introduction, first paragraph of

discussionFor each experiment

What is the specific hypothesis? end of the introductionWhat are the independent variables? in the figures, on the X-axis of graphs,

method, resultsWhat are the dependent variables?What’s measured?

in the figures, on the Y-axis of graphs,method, results

What are the most important results? first part of discussionWhat is the interpretation of most impor-tant results?

abstract, first and last paragraph of dis-cussion

Are there weaknesses, limitations? middle part of discussion, find additionalones yourself!

NotesRead the abstract thoroughly. Everything in there is important (but perhaps difficultto fully understand without reading more).Part of the discussion may be labelled 'Conclusion(s)'. If so, this will often be thesame as the last paragraph of the discussion in an article without a separate subhead-ing 'conclusion(s)'.In papers describing multiple experiments, the closing discussion or conclusion isgenerally referred to as General Discussion.

Applying this reading tool to one of my own studies renders the following overview:

General aspects of the paper

Question Place to lookWhat is the aim of the pa-per?

What question is ad-dressed?

abstract, end of the introduction, finalparagraph discussion;first paragraph of discussion

What experimental para- Introduction and method

50

digm is used? Has it beenused before, and what isthe typical interpretation?What underlying processesare being 'tapped' by theparadigm?

What is the general hy-pothesis?

end of the introduction, first paragraph of discussion

For each experimentWhat is the specific hy-pothesis?

end of the introduction

What are the independentvariables?

in the figures, on the X-axis of graphs, method, resultsExperiment 1:

Experiment 2:

What are the dependentvariables? What’s meas-ured?

in the figures, on the Y-axis of graphs, method, results

What are the most impor-tant results?

first part of discussion

51

What is the interpretationof most important results?

abstract, first and last paragraph of discussion

Are there weaknesses,limitations?

middle part of discussion, find additional ones yourself!

Finding the relevant pieces of information doesn’t yet make an adequate summary,but it definitely helps. Note that I was unable to answer all questions from the readingtool. It seems as though in my very own article I ‘forgot’ to formulate general andspecific hypotheses. I devised research questions but not hypotheses. That happenssometimes. Some researchers always provide a very clear hypothesis in their writings,others don’t. Don’t fret too much about it. Try to find the answers to all the questionsformulated above in the reading tool and if you can’t find the hypotheses you shouldshrug your shoulders, be mindful about it and then get to work. The ‘work’ of courserefers to stringing the information together into a coherent and eloquent summary.Writing an articulate summary means that you will have to go to the trouble to avoidpsychobabble. The danger otherwise is that you will use concepts or jargon words thatare so abstract (that’s what I mean by psychobabble) that your summary becomesunintelligible. What follows is an example of a summary I wrote largely with thesnippets of info I collected above with the reading tool.

52

Example

Written assignment 3 (PSY1136) SUMMARYStudent: Remco HavermansStudent ID: 13101974Article:Havermans, R. C., Geschwind, N., Filla, S., Nederkoorn, C., & Jansen, A. (2009).Sensory-specific satiety is unaffected by manipulations of flavour intensity. Physiol-ogy & Behavior, 97, 327-333.Word count: 369

The authors wanted to find out whether manipulating the intensity of a food or drinkwill affect the development and expression of sensory-specific satiety. To answer thisquestion, the authors devised two experiments. In each experiment, participants wereinstructed to first rate pleasantness of different food/drink items. Next, they receivedseveral servings of one of the foods/drinks (i.e., strawberry lemonade), after whichthey again had to evaluate all food items/drinks. It was expected that this would leadto decreased liking for the lemonade as compared to the other foods/drinks, indicativeof sensory-specific satiety.

In the first experiment, the research question was whether flavour intensity(weak versus strong) of the lemonade would affect sensory-specific satiety to it. In thesecond experiment, it was investigated whether varying the intensity of the lemonadefrom one serving to the next would attenuate sensory-specific satiety. For experiment1, the independent variables were Group (Strong lemonade versus Weak lemonade)and Food Type (the consumed lemonade, the unconsumed lemonade, or the controlfood/drink items). For the second experiment the groups were Increase (intensity in-creases with each serving), Same, or Decrease (intensity of the lemonade decreaseswith each serving), and Type again served as the other independent variable but nowthis factor had just two levels, comparing the ‘consumed’ (i.e., the strawberry lemon-ade) with the ‘unconsumed’ (the other foods/drinks). For both experiments, the maindependent variables were the changes (from pre- to post-test) in subjective ratings ofpleasantness and prospective consumption. By prospective consumption the authorsmean the rated desire or motivation to drink or eat more of each of the differentfoods/drinks.

Sensory-specific satiety for the lemonade was found in both experiments, butthe different flavour intensity manipulations in the two experiments did not affect thissatiety at all. The authors regard these results as supportive of the conjecture that sen-sory-specific satiety is reflected by neural habituation in particularly the orbitofrontalregion and not the insular cortex, further corroborating the notion that the intensityand reward value of a flavour are processed separately and independently. The au-thors however point out that the manipulation of the lemonade intensity comprised the

53

simultaneous manipulation of aroma (strawberry) and taste (sweetness) intensity. Thismight have confounded the results of their second experiment.

54

Skills IV (work in progress… )

55

Bachelor thesis

The bachelor thesis is a paper describing a literature or empirical study. Preferablybefore 24 December, but at least before the end of January the following year, youneed to have found a supervisor and have some idea of what your thesis will be about.Note that – as of 2010 – you NO LONGER are required to hand in an elaborate prob-lem statement. You may, of course, but it’s no longer required.

Write your thesis when and where you like, but note that you need to hand in afirst FULL version to your supervisor before the end of April, which obliges yoursupervisor to provide elaborate feedback in May. Revise your thesis according to thecomments by your supervisor and hand in your final bachelor thesis no later thanJune 15. Hand in a digital copy to: fdp-bachelorthese@maastrichtuniversity.nl. Andsubmit a neat paper copy of your thesis to Myrtle Brongers at the FPN Education Of-fice after the approval of your thesis by your supervisor. We need to have these docu-ments before the end of August.

Bachelor thesis guidelinesThe thesis needs to be a minimum of 6000 words (excluding title page, table of con-tents, reference list, notes and appendices) and a maximum of 10000 words. If yoursupervisor allows it, you can submit a slightly shorter version (e.g., 5743 words), butif your thesis will be substantially shorter than 6000 words, you and your supervisorneed to argue why the thesis cannot be 6000 words long and you are required to for-mally ask permission from the FPN Exam Committee for writing such a brief bache-lor thesis.

A thesis is in most cases a literature review but such reviews are also guided by aresearch question. As a literature review is research just as much as any experiment, itmakes senses to also include (at least) a paragraph on your methods for finding andselecting papers for your thesis. You should discuss the usefulness of adding such aparagraph (or perhaps even a whole section) with your supervisor. You are allowed toomit this paragraph only if your supervisor feels it does not really fit your thesis.

Importantly, use APA guidelines for referencing. No other referencing guidelinesare allowed, even when your supervisor would prefer to see it otherwise. The bachelorthesis is written in Dutch or English, not a mixture of both.

Bachelor thesis outline1. Title page: apart from the title, state your name, ID number, and the name of

your supervisor. Further indicate on the title page whether your thesis con-cerns EMPIRICAL RESEARCH or a LITERATURE REVIEW and provide atotal word count (excluding title page, table of contents, reference list, notesand appendices).

56

2. Table of contents (note that this is not usually the case for a research or reviewarticle, but given that the bachelor thesis comprises a larger writing assign-ment, much larger than a typical article, this table of content is considered use-ful)

3. Abstract4. Introduction5. Central sections (for empirical research these sections always include Methods

and Results)6. Discussion and/or Conclusions7. References8. Notes (if any) or appendices (if any)

( ! ) Please note that in the case you’re writing a bachelor thesis based on yourMARBLE research, the above deadlines may not apply to you. Please check with theMARBLE coordinator what deadlines you specifically need to adhere to.

57

Master thesis

The master thesis is a report of the research you carried out during your internship.Before you start your internship research, you need to complete a research proposal,or protocol. This research protocol is evaluated by both your first and second supervi-sor. Only if this protocol is approved by both supervisors/assessors you may start withyour internship research. Please note that the research proposal must be submittedwithin 4 weeks of commencing the research internship.

Research protocol guidelines

A complete research proposal consists of the following parts:• a brief theoretical background of the research• the question(s) addressed by the research• a description of the research plan• a description of the research methods to be applied• a description of the techniques to be used for the processing and analyzing of

data• a reference list• a timetable

As a rough guideline, the length of the research proposal should be between 8 and 10A-4 pages. You can of course use large parts of this research proposal for the intro-duction and method section of your master’s thesis.

Master thesis guidelines

The formal guidelines for the master thesis are as follows:• The length of the thesis must be around 8000 words long (1.5 line spacing,

font size 12 [Times New Roman; 11 pts for ARIAL], margins of about 2.5 cmall round), excluding title page, references, tables, figures and appendices. Ifyour supervisors allow it, you can submit a slightly shorter version (e.g., 7243words), but if your thesis will be substantially shorter than 8000 words, youand your supervisor need to argue why the thesis cannot be at least 8000words long and you are required to formally ask permission from the FPNExam Committee for writing a shorter master thesis.

• The thesis must be written in English. Either you write your thesis in BritishEnglish or in American English, not a mix of the two.

• The thesis must be written individually.• Use APA guidelines for referencing. No other referencing guidelines are al-

lowed.

58

Thesis assessmentTwo supervisors/assessors assess the Master Thesis. The main assessor is the personwho supervised the research as far as its content is concerned; this could be the fac-ulty supervisor or the (external) internship supervisor. In that case the faculty supervi-sor is the second assessor. The Master Thesis will be assessed on aspects of content aswell as format and must comply fully with the norms set for content, structure, form(layout) and language usage, as described in this handbook. Criteria related to content(formulation of the question, line of reasoning, specialized aspects) and form will beassessed separately and contribute to the final mark for the thesis (see also the pro-spectus). The final version of the Master Thesis must be submitted in four copies, ofwhich three are hard (paper) versions and one an electronic copy. The three hard cop-ies must be submitted to the Education Office of the Faculty of Psychology and Neu-roscience. The Education Office sends one copy to the first assessor and another tothe second assessor. The third copy is meant for the university library. You may alsoprovide your supervisors directly with their own hard copy of your thesis in whichcase you only need to hand in a single hard copy at the Education Office. Please notethat it is not acceptable when your hard copies are merely a stack of papers with asingle staple in them! These should be bound together neatly, not stapled or with aring binding (i.e., “ringband”).

The electronic version must be sent as an attachment to the following e-mailaddress: fdp-masterthese@maastrichtuniversity.nl. The e-mail you send has to containthe name of the assignment (in this case masterthesis) and your ID-Number. The filename of the attachment must also include the student ID-Number (i.e. ‘masterthesisIDxxxxx.doc). You thus submit your thesis in the following manner. You send an e-mail to: fdp-masterthese@maastrichtuniversity.nl. In the ‘subject’ box you fill in‘masterthesis IDxxxxx’. Please note that you must write ‘masterthesis’ as one word,followed by a space and then your ID-Number. Do not forget to send in your thesis asan attachment.

Six weeks before the graduation date you have to hand in your master thesis toyour supervisors (either directly or via Education Office, see above). Your supervisorshave to correct and evaluate your thesis within 20 working days (4 weeks). If yoursupervisors agree with less time to check your thesis (e.g., 1 week), then you are al-lowed to hand in your thesis (a week) later. Two weeks before you graduate you haveto hand in all your study results at the Education Office.

Master thesis outline1. Title page: apart from the title, state your name, ID number, master track, date,

location of internship, and the name and affiliation of each supervisor/assessor.Furthermore, provide a total word count (excluding title page, table of contents,reference list, notes and appendices).

59

2. Table of contents (note that this is unusual for a research article, but given thatthe master thesis comprises a larger writing assignment, much larger than a typi-cal article, this table of contents is considered useful)

3. Summary/Abstract4. Introduction5. Methods6. Results7. Discussion and/or Conclusions8. References9. Notes (if any) and/or appendices (if any)