Embed Size (px)
Citation preview
SIMM16GRADUATE SCHOOL, FACULTY OF SOCIAL SCIENCES, LUND UNIVERSITY
Introduction to Quantitative Methods
Spring 2019, version 1.0
For an always up-to-date online version, please visit the course site (check regularly for updates)
Welcome!
SIMM16
Welcome to the Autumn term’s course Introduction to Quantitative Methods
The aim of this course is for students with little prior knowledge of quantitative methods to develop an understanding of the basic concepts and fundamental principles guiding the use of quantitative methods, acquire basic practical skills with regard to the performance of statistical analysis and develop the ability to critically assess quantitative research. The participants formulate a research question that includes a hypothesized causal relationship and that can be addressed using an available dataset. During the course different techniques for processing and analyzing data will be introduced and the participants will, mainly under teacher supervision, work on answering their own research question using the tools presented to them in the lectures. Participants will also learn to assimilate and evaluate existing quantitative social science research as it is presented in scientific journals and/or reports.
The course will focus on 1) performing basic statistical analysis (of secondary data, using SPSS) and 2) comprehending and evaluating scientific papers based on quantitative methods (with focus on the participants own areas of research).
The course will be given in English. It does not presume any previous knowledge of statistics.
SECTION 1
About Lund University and Graduate School
Lund UniversityLund University seeks to be a world-class university that works to understand, explain and improve our world and the human condition. The University is ranked as one of the top 100 in the world. We tackle complex problems and global challenges and work to ensure that knowledge and innovations benefit society. We provide education and research in engineering, science, law, social sciences, economics and management, medicine, humanities, theology, fine art, music and drama.
Our 42,000 students and 7,400 employees are based at our campuses in Lund, Malmö and Helsingborg. The University has a turnover of around SEK 8,000 million (EUR 800 million), of which two thirds is in research and one third in education.
We are an international university with global recruitment. We cooperate with 600 partner universities in over 70 countries and are the only Swedish university to be a member of the strong international networks LERU (the League of European Research Universities) and Universitas 21.
2
CONTACT INFO
1. Graduate School Find us on a mape-mail: [email protected] page: graduateschool.sam.lu.se/Facebook: tinyurl.com/GS-Lund-facebookLinkedIn: tinyurl.com/GS-Lund-Linkedin
2. Student UnionHome page: https://www.samvetet.org
3. Lund UniversityHome page: www.lu.seThe university is on iTunesU, Youtube, Facebook, Instagram, LinkedIn and Twitter
The historic main university building, sometimes informally referred to as the White House, from 1882.
Graduate SchoolGraduate School at the Faculty of Social Sciences offers interdisciplinary programmes and courses at Master’s level. We coordinate three interdisciplinary programmes in Global Studies, Development Studies and Social Studies of Gender, and offer courses in theory of science, research methodology and in interdisciplinary themes.
Graduate School was founded in 2007 in response to the Faculty of Social Sciences’ need to offer interdisciplinary programmes in areas not covered by specific departments. There are currently about 150 programme students roughly distributed in groups of 20 to 40 students per programme. In addition, Graduate School has approximately 300 students taking our classes in theory of science and scientific methodology. We have students from every continent, all with a background in the social sciences. Our international teaching staff come from different departments at the Faculty of Social Sciences, as well as from outside the Faculty.
The Graduate School office is located on the 2nd floor in Gamla Kirurgen, Sandgatan 3.
Director of Studies Mikael Sundström [email protected].
Graduate School administration Email: [email protected]
Phone: +46 (0)46 222 48 83, +46 (0)46 222 31 30
Student Reception (open Monday 10-13 Wednesday 10-13)
Postal Address
Programme AdministratorGraduate SchoolFaculty of Social SciencesLund University, Box 117221 00 Lund, Sweden
3
The home of Graduate School used to be part of the city’s Old hospital – and its bed wards and operation theatres have duly been transformed into (rather less blood-spattered) lecture theatres and meeting rooms (our classrooms used to be bed wards). The old mortuary is in the basement (we’ve been thinking about
organising a Halloween tour one of these years… maybe it’s time?)
SECTION 2
Course overviewTeaching and ExaminationTeaching includes lectures, teacher assisted exercises in practical statistical analysis (computer lab work) and seminars. The course is teaching intensive and requires a high degree of participation.
Formal student learning outcomesKnowledge and understanding
Upon completion of the course, the student shall demonstrate an understanding of basic concepts and fundamental principles associated with quantitative methods.
Competence and skills
On completion of the course the students shall, independently and with proficiency, show ability to:
๏ Perform basic statistical analysis
๏ Demonstrate a working knowledge of SPSS
๏ Select the appropriate method, interpret the outcome and report the results
๏ Formulate and in an appropriate way examine a hypothesis about a causal relationship
Judgement and approach
๏ Assimilate and reflect on texts (reports or scientific papers) where the argument is based on basic statistical analysis in a knowledgeable, independent and theoretically informed way
๏ Critically and independently reflect on methodological aspects of such analysis
.
4
AssessmentThe learning outcomes will be examined through two papers, the first presenting results of applied analysis, and the second providing a critical assessment of quantitative research studies.
Three opportunities for examination are offered in conjunction with the course: a first examination and two re-examinations. Within a year of the conclusion of the course, two further re-examination opportunities on the same course content are to be offered. After this, further re-examination opportunities are offered but in accordance with the then current course syllabus.
Grades
Marking scale: Fail, E, D, C, B, A. The highest grade is A and the lowest passing grade is E. The grade for a non-passing result is Fail (U).
The student’s performance is assessed with reference to the learning outcomes of the course. For the grade of E the student must show acceptable results. For the grade of D the student must show satisfactory results. For the grade of C the student must show good results. For the grade of B the student must show very good results. For the grade of A the student must show excellent results. For the grade of Fail the student must have shown unacceptable results.
The grading is based 75% on the first paper (applied analysis) and 25% on the second paper (assessment of quantitative research.)
At the start of the course students are informed about the learning outcomes stated in the syllabus and about the grading scale and how it is applied in the course.
5
SECTION 3
Your teachersYour teaching team
Mimmi Barmark (course coordinator) is associate professor at the department of Sociology. She has conducted research within the fields of social determinants of health and illness, as well as school choice and school segregation. She has extensive experience of teaching quantitative methods to students at all levels.
Kadri Kuusk is a PhD candidate at the Department of Human Geography. Her research focuses on long-term regional economic growth. In the past two years she has been teaching quantitative methods and GIS on several courses at the Faculty. She also has extensive experience in policy evaluation and teaching evaluation methods to policy makers.
Stine Madsen is a PhD candidate at the Department of Human Geography. Departing from an empirical focus on plastics and sustainability, she explores issues related to environmental governance, knowledge and policy.
Staffan Edling is a PhD candidate at the Department of Sociology. Drawing on perspectives from science and technology studies, he is doing ethnographic research on the aggregation, presentation and dissemination of quantitative and qualitative data in political organisations. He has experience teaching quantitative methods on undergraduate and master's level courses.
6
1. CONTACT INFORMATION
2. Mimmi Barmark (course coordinator)[email protected]
3. Kadri [email protected]
4. Stine [email protected]
5. Staffan [email protected]
Kadri StineMimmi Staffan
Course Resources – Books
Aneshensel, Carol S. (2013). Theory-Based Data Analysis for the Social Sciences. (2nd ed.). London: SAGE. Aneshensel, Carol S. (2013). Theory-Based Data Analysis for the Social Sciences. (2nd ed.). London: SAGE.
From the blurb: This book presents a method for bringing data analysis and statistical technique into line with theory. The author begins by describing the elaboration model for analyzing the empirical association between variables. She then introduces a new concept into this model, the focal relationship. Building upon the focal relationship as the cornerstone for all subsequent analysis, two analytic strategies are developed to establish its internal validity: an exclusionary strategy to eliminate alternative explanations, and an inclusive strategy which looks at the interconnected set of relationships predicted by theory. Using real examples of social research, the author demonstrates the use of this approach for two common forms of analysis, multiple linear regression and logistic regression. Whether learning data analysis for the first time or adding new techniques to your repertoire, this book provides an excellent basis for theory-based data analysis.
From the blurb: This book presents a method for bringing data analysis and statistical technique into line with theory. The author begins by describing the elaboration model for analyzing the empirical association between variables. She then introduces a new concept into this model, the focal relationship. Building upon the focal relationship as the cornerstone for all subsequent analysis, two analytic strategies are developed to establish its internal validity: an exclusionary strategy to eliminate alternative explanations, and an inclusive strategy which looks at the interconnected set of relationships predicted by theory. Using real examples of social research, the author demonstrates the use of this approach for two common forms of analysis, multiple linear regression and logistic regression. Whether learning data analysis for the first time or adding new techniques to your repertoire, this book provides an excellent basis for theory-based data analysis.
472 pages ISBN 978-1-4129-9435-4
7
Muijs, Daniel (2010). Doing Quantitative Research in Education With SPSS. (2nd ed.). London: SAGE. Muijs, Daniel (2010). Doing Quantitative Research in Education With SPSS. (2nd ed.). London: SAGE.
From the blurb: Using datasets from real-life educational research and avoiding the use of mathematical formulae, the author guides students through the essential techniques that they will need to know, explaining each procedure using the latest version of SPSS. The datasets can also be downloaded from the book's website, enabling students to practice the techniques for themselves.
This revised and updated second edition now also includes more advanced methods such as log linear analysis, logistic regression, and canonical correlation.
Written specifically for those with no prior experience of quantitative research, this book is ideal for education students and researchers in this field.
From the blurb: Using datasets from real-life educational research and avoiding the use of mathematical formulae, the author guides students through the essential techniques that they will need to know, explaining each procedure using the latest version of SPSS. The datasets can also be downloaded from the book's website, enabling students to practice the techniques for themselves.
This revised and updated second edition now also includes more advanced methods such as log linear analysis, logistic regression, and canonical correlation.
Written specifically for those with no prior experience of quantitative research, this book is ideal for education students and researchers in this field.
247 pages ISBN 978-1-8492-0324-1
SECTION 4
Course ResourcesThe mandatory literature that will be presented in the next few pages, plus selected journal articles provided by course leaders, comprises some 664 pages.
Please read the detailed reading lists for each lecture and seminar (later in this guide) for the specific chapters you are asked to read after class.
There are also some web-based SPSS resources now available at Graduate School's webpage: https://graduateschool.sam.lu.se/education/information-about-your-studies-programme-students/student-links-and-resources/how-to-use-spss
SECTION 5
Detailed course overviewTeaching and Reading Schedule
8
Table 1 Your course at a glance
TIME COURSE ACTIVITY
25/3, 10-12.00 & 13-15.00 Lecture & Lab 1 | Kadri Kuusk & Stine Madsen/ Staffan Edling
26/3, 10-12.00 & 13-15.00 Lecture & Lab 2 | Kadri Kuusk & Stine Madsen/ Staffan Edling
28/3, 10-12.00 & 13-15.00 Lab 3-4 | Kadri Kuusk & Stine Madsen/ Staffan Edling
1/4, 10-12.00 & 13-15.00 Lecture & Lab 5 | Kadri Kuusk & Stine Madsen/ Staffan Edling
2/4, 10-12.00 & 13-15.00 Lab 6-7 | Kadri Kuusk & Stine Madsen/ Staffan Edling
3/4, 13-15.00 Lab 8 | Kadri Kuusk & Stine Madsen/ Staffan Edling
5/4, 10-12.00 & 13.15.00 Lecture & Lab 9 | Kadri Kuusk & Stine Madsen/ Staffan Edling
8/4, 13-15.00 Lab 10 | Kadri Kuusk & Stine Madsen/ Staffan Edling
10/4, 10-12.00 & 13-15.00 Lecture + Lab 11 | Kadri Kuusk & Stine Madsen/ Staffan Edling
11/4, 10-12.00 Lab 12 | Kadri Kuusk & Stine Madsen/ Staffan Edling
15/4, 10-12.00 & 13-15.00 Lecture + Lab 13 | Kadri Kuusk & Stine Madsen/ Staffan Edling
16/4, 10-12.00 & 13-15.00 Lab 14-15 | Kadri Kuusk & Stine Madsen/ Staffan Edling
17/4, 9-12.00 Seminar | Kadri Kuusk & Stine Madsen/ Staffan Edling
24/4, 13-16.00 Seminar | Kadri Kuusk & Stine Madsen/ Staffan Edling
26/4, 13-15.00 Supervision | Kadri Kuusk & Stine Madsen/ Staffan Edling
29/4, 16.30-17.00 Deadline - Assignments 1 & 2
NB. Regularly check the course lesson plan in Live@Lund for potential schedule alterations and to find your relevant classrooms
NB. Regularly check the course lesson plan in Live@Lund for potential schedule alterations and to find your relevant classrooms
Introduction (lecture) | Teachers: Kadri Kuusk & Stine Madsen
Introduction to the course. The key concept of the course, the "focal relationship", is introduced and we learn about the basic principles of quantitative analysis and statistical inference.
Primary reading
Aneshensel, Carol S. (2013) Chapters 1-2
Muijs, Daniel (2010) Chapters 1-5
Lab 1. Introducing SPSS and data sets (lab) | Teachers: Kadri Kuusk & Stine Madsen/ Staffan Edling
We familiarize ourselves with the statistical software (SPSS) and the datasets we will use in the course.
Measurement and descriptives (univariate analysis) (lecture) | Teachers: Kadri Kuusk & Stine Madsen
This lecture covers two topics: measuring social phenomena (including the construction of indices/scales) and descriptive (univariate) analysis. We will in relation to this also learn how to identify a variable's "level of measurement".
Primary reading
Muijs, Daniel (2010) Chapters 6 & 11 (p. 217-223)
Lab 2. Research question workshop(lab) | Teachers: Kadri Kuusk & Stine Madsen/ Staffan Edling
Students formulate the research question (involving a focal relationship) that they will answer through their work on the first assignment.
Lab 3. Index construction (lab) | Teachers: Kadri Kuusk & Stine Madsen/ Staffan Edling
We will learn how to construct indices.
Lab 4. Descriptive statistics(lab) | Teachers: Kadri Kuusk & Stine Madsen/ Staffan Edling
Students work on the first assignment.
9
Analyzing relationships (bivariate analysis) (lecture) | Teachers: Kadri Kuusk & Stine Madsen
Bivariate analysis I: crosstabs and comparison of means: In this lecture we will learn how to analyze relationships between two variables using crosstab analysis and comparison of means.
Bivariate analysis II: regression analysis: We will learn how to analyze a relationship between two continuous variables using linear regression analysis.
Primary reading
Aneshensel, Carol S. (2013) Chapter 3 (p. 57-69) & Chapter 5 (p. 125-142)
Muijs, Daniel (2010) Chapters 7-8
Lab 5-6. Analyzing relationships I(lab) | Teachers: Kadri Kuusk & Stine Madsen/ Staffan Edling
Analyzing relationships, part I: we will practice how to conduct and interpret a crosstab analysis and comparison of means in SPSS.
Lab 7-8. Analyzing relationships II (lab) | Teachers: Kadri Kuusk & Stine Madsen/ Staffan Edling
Analyzing relationships, part II: we will practice how to conduct and interpret a regression analysis in SPSS. We will also learn how to produce and interpret a scatterplot and the correlation coefficient Pearson's r.
Multivariate analysis: The third variable(lecture) | Teachers: Kadri Kuusk & Stine Madsen
We will learn how we can uncover more about our focal relationship by adding a third variable to the analysis. We focus on how to choose a relevant third variable and how to interpret the outcome through applying Aneshensel's elaboration model on multivariate contingency tables and comparison of means.
Primary reading
Aneshensel, Carol S. (2013) Chapter 3 (p. 69-81), & Chapter 4
Lab 9-10. Multivariate analysis(lab) | Teachers: Kadri Kuusk & Stine Madsen/ Staffan Edling
We will practice the above and work on individual assignments.
10
Multiple regression analysis I: The basics (lecture) | Teachers: Kadri Kuusk & Stine Madsen
We will learn how to conduct and interpret a multiple regression analysis, following Aneshensel's elaboration model.
Primary reading
Aneshensel, Carol S. (2013) Chapter 5 (p. 143-163) & Chapter 7-9
Muijs, Daniel (2010) Chapter 9 (p. 138-147)
Lab 11-12. Multiple regression analysis I(lab) | Teachers: Kadri Kuusk & Stine Madsen/ Staffan Edling
We will practice the above and work on individual assignments.
Multiple regression analysis II: Refining the model(lecture) | Teachers: Kadri Kuusk & Stine Madsen
We will learn more about the assumptions underlying linear regression analysis and how to examine non-linear relationships and interaction effects.
Primary reading
Aneshensel, Carol S. (2013) Chapter 11
Muijs, Daniel (2010) Chapter 9 (p. 147-157) & Chapter 10 (p. 175-187)
Lab 13. Multiple regression analysis II(lab) | Teachers: Kadri Kuusk & Stine Madsen/ Staffan Edling
We will practice constructing dummy variables and including interaction terms in regression. Finally, we will put together everything we have learned about regression, and work on our individual assignments.
11
Lab 14-15. Individual work(lab) | Teachers: Kadri Kuusk & Stine Madsen/ Staffan Edling
Work on individual assignments.
Seminar on assignment 2(seminar) | Teacher: Kadri Kuusk & Stine Madsen/ Staffan Edling
Students will get advice in relation to the second assignment. In the first part of the seminar students will be introduced to the ways and tools of analysing research design by using a practical example. In the second part, students will practice in groups how to critically read and assess quantitative research on their own.
Peer review seminar (Assignment 1)(seminar) | Teachers: Kadri Kuusk & Stine Madsen/ Staffan Edling
Students will present their preliminary findings (Assignment 1) in small groups and receive feedback from the other students and their teacher.
Tutorial(supervision) | Teacher: Kadri Kuusk & Stine Madsen/ Staffan Edling
Students have the opportunity to receive some final guidance on the assignments.
Hand-in
Students publish their assignment in the designated folders on Live@Lund at 5pm at the latest.
Re-examination
Students publish their assignment in the designated folders on Live@Lund at 5pm at the latest.
12
Assignment 1: Analyse secondary dataYou will be given a couple of databases to choose from. The task is to formulate a research question that includes a (possible) causal relationship between two phenomena and use the statistical methods provided in the course to answer that question.
Follow the steps below when analysing your focal relationship:
1. Construct an index
Select a few variables (preferably three or more) that you on theoretical grounds perceive of as indicators of the same underlying phenomenon – your latent dependent variable. Create an index/summated scale out of these variables. Motivate your choice of variables and test whether they are “internally consistent”. You can then treat this index as a quantitative (numerical) variable in the upcoming analysis.
2. Analyse your focal variables (univariate analysis)
Illustrate the distribution of your index variable with a graph and appropriate measures of central tendency and dispersion. Illustrate also your focal x-variable in an appropriate way. If you recode this variable, report both the original and the recoded version of the variable and explain why you chose to recode it in this way.
3. Analyse your focal relationship (bivariate analysis)
Estimate the strength and significance of your focal relationship using contingency tables, comparison of means (when appropriate) and regression analysis.
4. Analyse your focal relationship (multivariate analysis)
Use multivariate crosstab analysis and multiple regression analysis to further test and explore your focal relationship. Be clear about whether you are using an exclusionary and/or inclusive strategy. Make sure to analyse how the focal relationship (strength and significance) is affected by the inclusion of additional variables and what conclusions you draw from this change (or absence of change).
5: Summate and discuss your results
Make sure that the focal relationship is at the centre of the discussion. The discussion must be theoretically informed, i.e. include reasoning about possible causal mechanisms. References to theoretical literature and/or previous empirical studies are welcome but not mandatory.
Follow the steps above when reporting your results. Only include tables and graphs that you refer to in the text. Make sure to report your results in a “non-statistical” language, especially in the concluding discussion, so that they are understandable to someone who has not taken this course. The paper should comprise about 10-15 pages, including text and tables. Put all tables in running text (no attachments). Please save as a Word document so we can provide feedback directly in the document.
Assignment 2: Write a (short) research overviewTake your own research interest as a starting point and define a ”quantitative” research question that includes one focal relationship. This can be the same one you define in assignment 1, or something completely different.
Search for 3 articles (published in scientific journals) that in some way address this question. The articles should use the statistical techniques you are now familiar with. For students who prefer this, there will also be the option to read a set of articles provided by the teachers.
Summate very briefly the content of the articles, then compare and evaluate the (quantitative) methods used. Aspects of the studies that should be critically discussed are e.g. sampling strategy, missing data, operationalization, validity and analytic strategy (i.e. the relationship between theory/research question and statistical analysis). How do the authors go about exploring their focal relationship? How do they motivate their different methodological choices and do you buy the arguments?
Finally summate the conclusions that can be drawn from these studies and 1) discuss ways in which the results can contribute to/be integrated with qualitative research within this area and/or 2) based on the potential weaknesses/gaps you spot, make some suggestions for future research.
The lion’s share of the paper should consist of a discussion of the similarities and differences, strengths and weaknesses, of the papers (i.e. not just description).
Your paper should be readable as a standalone piece; i.e. do not use text like ‘see table 3 in paper 1’. Ensure you include a bibliography of the articles you read. Please add a link to these articles.
The paper should be between about 3-4 pages (or 8000 characters without spaces).
13
Non-attendance at lectures and seminarsExcept from the introduction, none of the teaching in this course (including lectures, laborations and seminars) is mandatory. Students are however highly recommended to participate to an as high degree as possible.
Re-examination opportunitiesIf you miss or fail your exam assignment you will be offered two more chances later in the term – dates and times will be communicated by the course leaders.
PlagiarismAll the work for this course involves learning to do methods in a way that shows transparency in the research process, and academic integrity in your own work. This means we place a high emphasis on responsibility and trust in the research process. Your work must be original and a result of your research skills and intellectual efforts.
We follow the Graduate School guidelines and this means the following: all sources must be referenced clearly. Cutting and pasting from the Internet (or elsewhere) is not permitted, although Internet sources can be cited if referenced. If you hand in an assignment that to a large part consist of material cut and pasted from the Internet or quote texts authored by others without references it is considered as intentional deception. Copying other students, helping someone to cheat or letting a fellow student copy your texts are other examples of intentional deception.
If you intentionally try to deceive the examiner you will be reported to Lund University’s Disciplinary Board and will be subjected to disciplinary action.
14
