Squire’s Quest!

Squire’s Quest!Dietary Outcome Evaluation of a Multimedia GameTom Baranowski, PhD, Janice Baranowski, MPH, RD, LD, Karen W. Cullen, DrPH, RD, LD,Tara Marsh, MS, RD, Noemi Islam, MS, RD, Issa Zakeri, PhD, Lauren Honess-Morreale, MPH,Carl deMoor, PhD

Background: Fruit, juice, and vegetable (FJV) consumption among children is low. Innovative programsare needed to enable children to increase FJV intake. Psychoeducational multimediapermits the delivery of interventions as designed and capitalizes on known behavior changeprinciples.

Design: Elementary school was the unit of recruitment, assignment, and analysis. Twenty-sixelementary schools were pair matched on size and percentage of free or reduced-pricelunch, and randomly assigned to treatment or control groups. Data were collected justbefore and just after the program.

Setting/Participants:

All fourth-grade students in participating elementary schools were invited to participate.Data were collected on 1578 students.

MainOutcome:

Servings of fruit, 100% juice, and vegetables consumed.

Intervention: Squire’s Quest! is a ten-session, psychoeducational, multimedia game delivered over 5 weeks,with each session lasting about 25 minutes. Based on social cognitive theory, educationalactivities attempted to increase preferences for FJV through multiple exposures and associatingfun with their consumption, increase asking behaviors for FJV at home and while eating out,and increase skills in FJV preparation through making virtual recipes.

Measures: Four days of dietary intake were assessed before and after the intervention. Assessment wasmade by the Food Intake Recording Software System (FIRSSt), which conducts a multiplepass, 24-hour dietary intake interview directly with the children.

Results: Children participating in Squire’s Quest! increased their FJV consumption by 1.0 servingsmore than the children not receiving the program.

Conclusions: Psychoeducational multimedia games have the potential to substantially change dietarybehavior. More research is warranted. (Am J Prev Med 2003;24(1):52–61) © 2003American Journal of Preventive Medicine

Background

People who consume more fruit, 100% juice, andvegetables (FJV) have greater longevity1 andsome level of protection from several cancers,2

heart disease,3 diabetes mellitus,4 and perhaps evenaging of skin.5 While children do not ordinarily expe-rience adult chronic diseases, some cancers have a longdevelopmental period, perhaps initiating at puberty.6

Furthermore, food-related preferences and practicesstart in the earliest years,7 and FJV consumption maytrack (i.e., those at higher levels of FJV consumption atyounger ages remain higher consumers later inlife).8–10 Thus, interventions to promote higher FJVconsumption among children hold the promise ofimmediate healthier growth for children11; the preven-tion of the initiation of cancer during adolescence, ifcarried through puberty6; and the prevention of cancerand other illnesses in mid-life, if carried into adult-hood.

Nutrition education programs effect change in eat-ing FJV through changing mediating variables.12,13 Themost important mediators of eating more FJV includeincreasing the availability and accessibility of FJV athome and when eating out,14 increasing FJV preferenc-es,7,15 and increasing children’s skills at making FJV

From the Children’s Nutrition Research Center, Department ofPediatrics, Baylor College of Medicine (T Baranowski, J Baranowski,Cullen, Marsh, Islam, Zakeri), and M.D. Anderson Cancer Center,Department of Behavioral Science, University of Texas (Honess-Morreale, deMoor), Houston, Texas

Address correspondence and reprint requests to: Tom Baranowski,PhD, Professor of Pediatrics (Behavioral Nutrition), The Children’sNutrition Research Center, Department of Pediatrics, Baylor Collegeof Medicine, 1100 Bates Street, Room 2038, Houston TX 77030-2600.E-mail: [email protected].

52 Am J Prev Med 2003;24(1) 0749-3797/03/$–see front matter© 2003 American Journal of Preventive Medicine • Published by Elsevier Science Inc. PII S0749-3797(02)00570-6

recipes when they are responsible for making their ownsnacks.16

Schools present an important channel for reachinglarge numbers of children with nutrition education.Only one school nutrition education program, whichwas implemented by specially trained teachers, resultedin substantial FJV change (1.6 servings) after 2 years ofprogram delivery.17 Others have achieved modest18,19

or no FJV change.20 One of these interventions re-vealed that the usual classroom teacher implementedonly 50% of the curriculum-specified activities withonly 22% of the activities likely to result in behaviorchange.21 Thus, channels need to be found that delivernutrition education programs more directly to children.

One channel that interacts with children directly iscomputer-based, interactive multimedia education(IMME).22 IMME is an attractive educational modalitybecause it can combine visual, aural, and text-basedmessages22 and incorporate entertainment into educa-tion (edutainment), thereby making the messagesmore acceptable and the activities more enjoyable.When based on known psychological principles, IMMEhas been called psychoeducational multimedia training(PEMT).23 PEMT can deliver the intervention exactlyas designed by the investigators, especially the precisespecification of theoretically prescribed, behaviorchange procedures.

Two early PEMT programs were targeted at high-riskbehaviors among adolescents.24,25 More recent PEMTprograms addressed physical activity,26 obesity,27,28 andeating disorders.29 A PEMT kiosk whose program wasbased on social cognitive theory in five grocery storesled to improved consumption of dietary fat, fiber, fruit,and vegetables.30 Adults receiving a “talking computer”intervention over the telephone increased fruit con-sumption by 1.1 servings per day.31 Computer-tailored,nutrition-education letters sent to families of adoles-cents resulted in reduced dietary fat intake32 among allfamily members; these letters worked, however, onlywith the mothers, not the fathers or adolescents.32 Noresearch on PEMT to promote dietary change has beenreported with elementary school children.

This paper reports the dietary outcome evaluation ofthe Squire’s Quest! PEMT nutrition education gamewith fourth-grade students in Houston, Texas.

MethodsResearch Design

Our objective was to demonstrate dietary change immediatelyafter implementation of the Squire’s Quest! program. Webelieved that it was important to demonstrate change with anew technology right after the intervention. To achieve thisobjective, a simple two-group design (treatment and control)with pre- and post-assessment was employed. Due to theeffects of data clustering, the school was the unit of recruit-ment, random assignment to group, and analysis. Baseline

assessment took approximately 2 weeks per school, the inter-vention was conducted in 5 weeks, and post-assessment tookapproximately 2 weeks per school. Thus, a cycle could becompleted in one school semester. Schools were matched onsize33 and percentage of students receiving free and reduced-price lunch (an indicator of socioeconomic status [SES]).Within matched pairs, schools were randomly assigned toconditions.

Power calculations revealed that 26 elementary schoolswere necessary to detect a 0.5 standard-deviation change inservings of FJV with 80% power and 5% type-one errorassuming a school-associated intraclass correlation of 0.02.The design was implemented in two waves. Initially, 14elementary schools were recruited and randomly assigned togroups in fall 1999. One control school declined to partici-pate after baseline assessment. As a result, 13 elementaryschools were recruited and randomly assigned to groups inspring of 2000.

Due to the limited number of computers for baselineassessment and intervention (n�76), and the need to leavethe computers in the treatment schools to run Squire’s Quest!(the variability in type and processing speed of computersalready in the schools was too large to anticipate in theprogramming), the treatment and control schools were iden-tified prior to baseline assessment. Baseline assessment wasconducted for approximately 2 weeks in the control schools,and then computers were moved to the treatment schools forassessment and intervention. The treatment group completedpost-assessment first, after which the computers were movedto the control schools.

Sample

The Houston Independent School District (the third largestschool district in the United States) agreed to participate. Acomprehensive recruitment program was implemented, start-ing with contact with administrators and presentations toprincipals, and followed by presentations to teachers atschools with principals agreeing for their schools to partici-pate. Decisions to participate were often made by the princi-pal in consultation with the teachers. Details of the recruit-ment procedures and the factors influencing schoolparticipation have been presented elsewhere.34 All childrencompleted an informed assent form, and parents completedan informed consent form for their child and themselves,which was returned by the child to school. The consents wereworded to indicate that the child could receive either condi-tion, but measurement was required in both.

In the treatment schools, 73.2% of students providedinformed consent to participate, and 67.6% of studentsprovided informed consent in the control schools. A total of1578 students were recruited to participate in Squire’s Quest!Characteristics of the children in the 26 schools that com-pleted post-assessments, as well as those who did not, appearin Table 1.

Squire’s Quest!

Squire’s Quest! was designed as a ten-session, interactivemultimedia game, with each session taking about 25 minutesto complete. The story line for the game was as follows: thekingdom of 5A Lot was being invaded by the Slimes (snakes)

Am J Prev Med 2003;24(1) 53

and the Mogs (moles), who were attempting to destroy thekingdom by destroying the fruit and vegetable crops. KingCornwell and Queen Nutritia were leading their knights (e.g.,Sir Sarah See-a-Solution and Sir Alex Try-to-be-Right) todefeat the invaders. In the first session, the fourth-grade childcommitted to becoming a squire in the pursuit of becominga knight to help the king and queen. The squire had to facechallenges in his/her quest. The challenges involved skillsand goals related to eating more fruit, 100% fruit juice, andvegetables. The squire prepared FJV recipes (in a virtualkitchen) to provide energy for the king and court to fight theinvaders. A wizard mentored the child through the chal-lenges, and the castle robot (Mad Maxie) facilitated many ofthe educational sessions. The invaders kidnapped the goodchef (Chef Karat) and replaced him with Chef Mog, a

bumbling fool, who always usurped the squire’s accomplish-ments to make himself look good to the king.

Before the end of each session, the child set goals to makethe recipe (prepared in the virtual kitchen) during thatsession, eat another FJV serving at a meal or as a snack, or toask for his/her favorite FJV to be more available at home. Thechildren participated in a decision-making activity betweentheir favorite fruit, juice, or vegetable and a more commonsnack. The FJV was selected based on the child’s foodpreferences reported at baseline. The decision criteria werethe three most important outcome expectancies reported bythe child at baseline.

Sessions 2 to 10 began with an assessment of whether thegoal from the previous session was completed, for whichdragon-scale points were assigned. A problem-solving routine

Table 1. Sample characteristics and assessment of differences at baseline

Sample characteristicsTotal sampleN/n (%)

Completedpre- andpost-assessmentn (%)

Completedpre-assessmentalonen (%)

Treatmentgroupn (%)

Controlgroupn (%)

Totals 1578 (100) 1489 (94.7) 89 (5.3) 749 (100) 740 (100)Group

Treatment 785 749 (50.3) 36 (40.4) 749 (100) —Control 793 740 (49.7) 53 (59.6 ) — 740 (100)

Agea (in years)8 32 30 (2.3) 2 (2.7) 26 (4.1) 4 (0.6)9 872 833 (63.3) 39 (52.0) 413 (65.8) 420 (61.0)10 405 383 (29.1) 22 (29.3) 159 (25.3) 224 (32.6)11 70 61 (4.6) 9 (12.0) 28 (4.5) 33 (4.8)12 12 9 (0.7) 3 (4.0) 2 (0.3) 7 (1.0)

GenderBoys 736 689 (47.5) 47 (54.0) 344 (47.0) 345 (46.9)Girls 803 763 (52.5) 40 (46.0) 373 (52.0) 390 (53.1)

Ethnic groupb

African American 268 261 (17.9) 7 (8.6) 140 (18.9) 121 (16.9)Euro-American 690 643 (44.1) 47 (58.0) 326 (44.0) 317 (44.2)Hispanic 476 452 (31.0) 24 (29.6) 229 (30.9) 223 (31.1)Other 105 102 (7.0) 3 (3.7) 46 (6.2) 56 (7.8)

Dietary intakec (in servings) Mean (�SD) Mean (�SD) Mean (�SD) Mean (�SD) Mean (�SD)F 1.7 (2.1) 1.7 (2.1) 1.7 (2.0) 1.8 (2.2) 1.6 (2.0)J 0.8 (1.5) 0.8 (1.5) 0.9 (2.3) 0.8 (1.5) 0.9 (1.5)RV 0.8 (1.2) 0.8 (1.2) 0.6 (0.9) 0.8 (1.3) 0.8 (1.1)FJV 3.4 (3.4) 3.4 (3.4) 3.2 (3.6) 3.5 (3.7) 3.2 (3.1)HFV 0.3 (0.6) 0.3 (0.6) 0.5 (0.9) 0.3 (0.7) 0.3 (0.5)Total FJV�HFV 3.6 (3.5) 3.6 (3.5) 3.7 (3.7) 3.8 (3.8) 3.5 (3.2)

Dietary intake

Control Intervention

Completedpre and post(n�749)

Completedpre only(n�36)

Completedpre and post(n�740)

Completedpre only(n�53)

F 1.8 (2.2) 1.8 (1.8) 1.6 (2.0) 1.7 (2.1)J 0.8 (1.6) 0.6 (0.7) 0.8 (1.3) 1.1 (3.0)RV 0.9 (1.3) 0.7 (1.0) 0.8 (1.1) 0.6 (0.9)Total FJV 3.5 (3.7) 3.0 (2.5) 3.2 (3.0) 3.4 (4.2)HFV 0.3 (0.7) 0.4 (0.6) 0.3 (0.5) 0.5 (1.0)Total FJV�HFV 3.8 (3.8) 3.4 (2.6) 3.5 (3.1) 3.9 (4.2)

Note: n varies by variable due to missing data.aChi-square�18.185, df�4, p�0.001 for pre- and post-assessment versus pre-assessment only comparison. Chi-square�27.733, df�4, p�0.001 forcontrol versus intervention comparison.bChi-square�8.362, df�3, p�0.039 for pre- and post-assessment versus pre-assessment only comparison.ct�2.18, df�1487, p�0.029 for control versus intervention fruit comparison, and t�2.69, df�1576, p�0.007 for control versus intervention HFVcomparison.F, fruit; FJV, fruit, juice, and vegetables; HFV, high-fat vegetables; J, 100% juice; RV, regular vegetables.

54 American Journal of Preventive Medicine, Volume 24, Number 1

Table 2. Squire’s Quest! game overview of sessions 1 through 10

Session Day Meal Skill Knowledge RecipeGoal setting/assignment Schema Miscellaneous

1 M-W After-school snack Decision making What counts as F or J? J 1 svg of F/J None Week 1: parentnewsletter

F vs non-F What counts as J svg? Drink 2 days at snackHow many svgs/day? 1 recipe Make and try

recipe, askingredients

2 T-S Breakfast Asking skills What counts as F? F 1 svg of F/J None Why eat breakfast?What counts as F svg? Drink 2 days at breakfast

or snackProblem-solving routineif student does notcomplete goal session 1

2 recipes Make and tryrecipe, askingredients

Needs to be a routinethat appears when goalhas not been met

3 M-W School lunch Problem solving What foods count as V? No Eat svg of V None Week 2: parentnewsletter

Decision making: Vvs other snack

What counts svg of V? Recipes 2 days at schoollunch

School lunch: whatcounts?

How many svgs/day? If Session 1 goal notdone, no make-up

4 T-S Dinner V Selecting F/V How to buy fresh F? Side 1 V svg 2 days atdinner

None Use senses for guidelines

Side dish Storage F/V How to buy fresh V? Dish Make and tryrecipe, askingredients

If Session 2 goal notdone, no make-up

Trash It game How to store fresh? 3 recipesStash It game How to buy canned,

frozen?5 M-W Afternoon Problem solving

specific to askingskill: asking withnegotiating

How to substitute V?Substitution dependson time and preference

Snack V svg 2 days atsnack

Introduce5-a-day plan

Week 3: parentnewsletter

Snacks 3 recipes Make and tryrecipe, askingredients

Knight eats2 at snack

Asking skill: Did youask? (Yes/No)

Yes: What happened?Got itemReinforcing message:Did not get item; list ofmost common problemswith solution (no time,not available)No: Why not? (forgot,not comfortable)If Session 3 goal notdone, no make-up

(continued on next page)

Am

JPrev

Med

2003;24(1)55

Table 2. (continued)

Session Day Meal Skill Knowledge RecipeGoal setting/assignment Schema Miscellaneous

6 T-S FF lunch ordinner

Decision making atFF restaurant

What V choices at FF? No 1 svg F/V at FF 5-a-day plan If Session 4 goal notdone, no make-up

Choosing F/V What F choices at FF? Recipes Restaurantalternatives forstudent; no FF

King eats 2F atbreakfast

How to make a F/Vsvg?

7 M-W CS after school/noon snack

Find F/V at CS What V choices at CS? No Serving of F/V atCS

5-a-day plan Week 4: parentnewsletter

Find it and buy itgame

What F choices at CS? Recipes Alternatives for noCS

Queen has2 at lunch

Store will have J asreplacement for soda

How ads manipulate What to look out for? If Session 5 goal notdone, no make-up

8 T-S V with lunch anddinner

Knowledge FJ and Vreview game

Review F/V knowledge Combo Eat V svg at lunchand dinner for 2days

5-a-day plan If Session 6 goal notdone, no make-up

Why try new F/V? Tastebuds change

Foods Wizard has2 at dinner

Sweetness of F

3 recipes9 M-W Desserts Decision making: F

vs other dessertReview F/V goals metand how they can berelated to developingschema

Dessert Try eating 5-a-dayschema

Studentcreatesschema

Week 5: parentnewsletter

3 recipes Make and tryrecipe, askingredients

If Session 7 goal notdone, no make-up

Outcome expectations:Good thing that happenwhen eat F/VFamily Feud–type show:“Survey says . . . Stronger,more energy, better eyesight”

10 T-S Party Schema Fun F/V at parties Partyfoods

Make goal to follownew schema to get5-a-day everyday!

Schemacheck;correct ifneeded

If session 8–9 not done,no make-up

Woven into recipepreparation

Knighting ceremony!

3 recipes Presentschema toking

Goal: Choose 2 behaviorchanges from programto continue

CS, convenience store; F, fruit; FF, fast food; J, 100% juice; svg, serving; V, vegetable.

56A

merican

Journal

ofPreven

tiveM

edicine,

Volum

e24,

Num

ber1

was employed to help the child assess how he/she mightchange practices to increase the likelihood of goal attain-ment. Table 2 presents a list of session activities, in roughorder of sequence within a session.

All children attained knighthood (ten possible levels) atthe end of the ten sessions, with the level determined by thenumber of dragon-scale points earned. Points were earnedprimarily by attainment of goals, with smaller amounts ofpoints earned from the educational games. Examples of thesegames included identifying what counted as fruit, whatcounted as vegetables, and whether a demonstration of askingwould likely result in making FJV more available.

Focus group discussions with fourth-grade children wereemployed in the design of Squire’s Quest! to assess interest inthe story line and to identify child-desired characteristics ofcharacters.35 Squire’s Quest! focused on fourth-grade chil-dren alone, based on advice from developers of children’sgames. The primary creative writer attended two nationalconferences of developers of interactive multimedia games tobenefit from their insights.

Measures

The primary outcome measure was servings of fruit, 100%juice, and vegetable consumption as assessed by the FoodIntake Recording Software System (FIRSSt). FIRSSt was ad-ministered for 4 days at baseline and at post-assessment. Weattempted to obtain 4 nonconsecutive days per child over a2-week period, but in some cases we obtained consecutivedays. FIRSSt is an interactive, multimedia dietary assessmentprogram that simulates a multiple-pass, 24-hour dietary recall(24hDR),36 completed by the child. FIRSSt was demonstratedto perform with 46% matches against observation of previousday’s school lunch and 60% matches against a dietitian-conducted 24hDR.36 This was only somewhat lower than thedietitian-conducted 24hDR against the observation of previ-ous day’s school lunch (59%).36

Demographic characteristics were assessed using standardquestions on the consent form sent home to the parents.

Statistical Analyses

Differences in categoric variables between those completingpre- and post-assessments and those completing pre-assess-ment alone, and between treatment and control groups, weretested by chi-square statistics. Differences between thesegroups in mean servings of FJV were tested by independent ttest. Analysis of variance was used to calculate intraclasscorrelations for the clustering effect of school and for reli-ability across 4 days. Independent sample t tests were used totest for differences between groups in consumption of FJV atbaseline. Mixed-model analysis of covariance was used to testfor differences between treatment and control groups (fixedeffect) in FJV and each FJV component at post-assessment,controlling for corresponding pre-intervention values and forthe clustering effect of school (variable effect). Matching wasnot included in these analyses.37–39 For methodologic rigor,these analyses were repeated using the mean for each schoolin a pair-matched t test, a weighted pair-matched t test, and aWilcoxon-signed ranks test. To test for possible moderation oftreatment effects by age, gender, or ethnicity, the demo-graphic term and a group � demographic interaction term

were added to the above in separate models. To control fornon-normal distributions of the dietary variables, these anal-yses were repeated with log-transformed data. The sample wasdivided into quartiles of fruit (F); 100% juice (J); regularvegetables (RV); total fruit, juice, and vegetables (TFJV);high-fat vegetables (HFV); and total fruit, juice, and vegeta-bles, with high-fat vegetables (TFJV�HFV) consumption atbaseline. Within quartiles, mean intake was calculated atbaseline and at post-assessment separately for treatment andcontrol groups.

Results

Only 5.3% of students did not complete pre- andpost-assessments. The modal category for age was 9years, with children completing pre- and post-assess-ment being slightly younger. Euro-American childrenwere somewhat less likely than other groups to com-plete both pre- and post-assessments. There were nodifferences between participation groups in gender orservings of FJV consumed at baseline (Table 1).

Despite random assignment of schools to condition,the children in the control group were slightly older,but there were no differences between groups bygender or ethnic group (Table 1). Children assigned tothe treatment group consumed slightly more fruit andhigh-fat vegetables at baseline.

At baseline, the intraclass correlation associated withthe clustering effect by school was 0.02 for F, 0.01 for J,0.02 for RV, and 0.03 for TFJV. The intraclass correla-tions for change from pre- to post-assessment were 0.03for F, 0.04 for J, 0.01 for RV, 0.05 for TFJV, 0.02 forHFV, and 0.05 for TFJV�HFV. The reliability-relatedintraclass correlations across 4 days of assessment weremodest and comparable between pre- and post-assess-ments (Table 3).

Children in the spring implementation wave weresomewhat less likely to complete all ten sessions (Table4). Although principals agreed to participate, we wereasked to leave several schools early in the spring, whenwe came within 2 weeks of the administration ofstandardized tests, to enable teachers to concentrate onteaching to the standardized test.

Using a mixed-model analysis, the difference inmeans between treatment and control groups at post-assessment, after controlling for pre-assessment values,was 0.91 servings of FJV or 1.0 servings of TFJV�HFVper day (Table 5). Contributions to this difference were

Table 3. Reliability intraclass correlations across 4 days ofassessment at pre- and post-assessment

Dietary intake Pre-assessment Post-assessment

Fruit 0.44 0.52100% juice 0.38 0.44Vegetable 0.38 0.46Total fruit, juice,

and vegetable0.59 0.58

Am J Prev Med 2003;24(1) 57

made from each food group, but the statistically signif-icant differences were for F (0.52 servings) and RV(0.24 servings) (Table 5). This analysis was repeatedwith a t test on the mean of each school weighted bynumber of students per school, a matched-pairs t test(not weighted), and a Wilcoxon-signed ranks test onmatched school means. These analyses revealed similarresults (Table 5). The analysis was also repeated delet-ing the one school that was not randomly assigned withthe same results. There was no evidence for moderationof this effect by age, gender, or ethnicity. Girls ate more(main effect) fruit and vegetables, but not juice. Nei-ther age nor ethnicity was related to any component ofFJV consumption. The same pattern of findings wasfound with log-transformed data, with the lone excep-tion that the outcome effect for vegetables was reducedto marginal significance (p�0.06).

To describe the process of dietary change, FJV con-sumption is consumption group at baseline with meanvalues at baseline and post-program for treatment andcontrol groups separately (Table 6). Substantial regres-

sion to the mean occurred in all food groups. In thethird quartile, however, the mean increased for thetreatment group and did not increase or declined inthe control group.

Discussion

The Squire’s Quest! PEMT game resulted in a 1.0serving difference of FJV between treatment and con-trol groups at the end of the 5-week, ten-session pro-gram, after controlling for baseline FJV consumption.The strengths of this research include the large sampleof schools and students, the mixed ethnic and SEScomposition of the sample, using the school as the unitof assignment and analysis, random assignment ofschool to condition, minimal differences in consump-tion by groups at baseline, the ability of the computerto deliver the intervention as designed, and similarresults for all approaches to outcome analysis. It is notclear, however, how long this increased change wouldbe maintained beyond the end of the program. The

Table 4. Number of Squire’s Quest! sessions completed by students

Number ofsessions

Fall 1999n (%)

Spring 2000n (%) Total n (%)

1 10 (2.4%) 1 (0.3%) 11 (1.4%)2 7 (1.7%) 2 (0.5%) 9 (1.1%)3 3 (0.7%) 2 (0.5%) 5 (0.6%)4 6 (1.4%) 1 (0.3%) 7 (0.9%)5 4 (1.0%) 7 (1.8%) 11 (1.4%)6 2 (0.5%) 11 (2.9%) 13 (1.6%)7 5 (1.2%) 19 (5.0%) 24 (3.0%)8 8 (1.9%) 43 (11.2%) 51 (6.4%)9 10 (2.4%) 71 (18.5%) 81 (10.1%)10 364 (86.9%) 226 (59.0%) 590 (73.6%)Total 419 (100%) 383 (100%) 802 (100%)

Table 5. Differences in food group intakes between treatment and control groups at post-assessment from Squire’s Quest!using four analyses

Dietary intake

Mixed-modelsanalysis ofvariance

Weightedmatched pairst test(n1*n2)/(n1�n2)

Matched pairst test(not weighted)

Wilcoxon-signedranks test onmatched schoolmeans

Fruit 0.52 0.70 0.66 NAF�9.47, p�0.002 F�7.64, p�0.006 t�3.458, p�0.005 Z� �2.76, p�0.006

100% juice 0.17 0.17 0.19 NAF�2.02, p�0.156 F�2.02, p�0.156 t�1.761, p�0.104 Z� �1.43, p�0.15

Regular 0.24 0.27 0.24 NAvegetables F�10.6, p�0.001 F�8.36, p�0.004 t�2.70, p�0.019 Z�2.27, p�0.023

Total fruit, juice, 0.91 1.14 1.09 NAand vegetables F�9.4, p�0.002 F�4.41, p�0.036 t�3.63, p�0.003 Z� �2.83, p�0.005

High-fat 0.09 0.12 0.11 NAvegetables F�2.6, p�0.107 F�8.52, p�0.004 t�2.11, p�0.057 Z� �1.85, p�0.064

Total fruit, juice, 1.01 1.26 1.20 NAand vegetables F�11.7, p�0.0007 F�4.24, p�0.040 t�4.127, p�0.001 Z� �2.90, p�0.004with high-fatvegetables

NA, not applicable.


average consumption of FJV at baseline was comparableto that reported elsewhere in the literature amongchildren.19,40 While the increase of 1.0 FJV servings inconsumption was substantial, it was not enough toachieve the goal of an average of five servings per day.This reinforces the idea that children need to beexposed to the five-per-day message from multiplechannels, and the messages likely need to be repeatedseveral times throughout childhood in developmentallyappropriate ways.

The change in FJV consumption in this study issecond largest in the literature after a serving differ-ence of 1.6 for the High 5 Alabama program.17 Thehallmark of both programs is that they were not imple-mented by the usual classroom teacher, who was docu-mented to implement only about half the activities incorresponding curricula.41 The substantial changes at-tained suggest that the social-cognitive theory frame-work on which both interventions were predicatedrepresent important aspects of behaviors and, thereby,provide a strong foundation for the design of dietarychange programs for children. PEMT programs basedon other theories would be useful to assess the extent towhich the theory or medium accounts for outcomes.

The pattern of change from pre- to post-assessmentby quartile (Table 6) was similar to other interven-tions.19 There was less of a decline in the highestquartiles of the treatment versus control groups andmore of an increase in the lowest quartiles. Thesesubstantial changes by quartile occurred despite 4 daysof pre- and post-assessment that should provide morestable estimates of intake. Thus, while a shift of about1.0 servings occurred in the mean, this was not simplythe addition of a serving to all cases. This patternsuggests that messages should be tailored based onbaseline consumption: “minimize decline” in thehigher groups and “try it” in the lower groups (where�25% of the sample were consuming no 100% juice orvegetables). This pattern likely explains the number ofgoals � baseline consumption–interaction effect re-ported elsewhere.42

The results reveal that psychoeducational multime-dia games can result in dietary behavior change. Itappears that the fun aspect kept the attention of thestudents and may have facilitated the change.43 It is notclear to what extent, or how, each of the followingcontributed to outcome: an interesting story line aboutaiding a king to fend off invaders, the interactivity of

Table 6. Mean baseline and post-assessment value for treatment and control by overall quartile at baseline

Dietaryintake

Baselinequartiles

Baseline Treatment Control

M (SD) n M (SD) n M (SD) n

FruitI 0.06 (0.10) 396 1.06 (1.87) 184 0.54 (1.06) 188II 0.68 (0.21) 394 1.31 (1.63) 196 0.79 (1.19) 175III 1.56 (0.34) 399 1.85 (1.90) 197 1.17 (1.34) 183IV 4.53 (2.54) 389 3.22 (3.15) 163 2.72 (2.63) 203

Juicea

I, II 0.07 (0.12) 791 0.57 (0.94) 360 0.57 (0.97) 384III 0.71 (0.21) 399 0.90 (1.17) 204 0.70 (0.98) 173IV 2.57 (2.26) 388 1.71 (2.35) 176 1.26 (1.48) 192

Regular vegetablesa,b

I, II 0.04 (0.08) 777 0.60 (0.96) 349 0.48 (0.82) 378III 0.69 (0.25) 412 1.04 (1.24) 212 0.70 (0.90) 182IV 2.49 (1.43) 389 1.76 (1.57) 179 1.37 (1.49) 188

Total fruit, juice, and vegetablesI 0.50 (0.37) 399 1.93 (1.91) 171 1.52 (1.65) 202II 1.73 (0.36) 390 2.88 (2.70) 191 2.11 (2.08) 181III 3.31 (0.56) 393 3.96 (3.30) 202 3.16 (2.49) 168IV 7.87 (3.83) 396 6.21 (4.43) 176 4.81 (3.76) 198

High-fat vegetablesa,b

I, II, III 0.04 (0.10) 1231 0.35 (0.70) 574 0.27 (0.49) 590IV 1.18 (0.76) 347 0.59 (0.78) 165 0.66 (0.89) 159

Total fruit, juice, and vegetables with high-fat vegetablesI 0.61 (0.41) 397 2.48 (2.55) 170 1.76 (1.69) 203II 1.98 (0.40) 392 3.08 (2.61) 200 2.59 (2.37) 172III 3.67 (0.63) 397 4.65 (3.71) 203 3.35 (2.42) 174IV 8.36 (3.84) 392 6.59 (4.66) 167 5.24 (3.91) 200

Notes: Values removed were post-assessment regular vegetables (17.13), and post-assessment high-fat vegetables (15.00). New maximum valueswere post-assessment regular vegetables (7.75), and post-assessment high-fat vegetables (5.38).aAt baseline, �25% of students had no juice intake, �25% of students had no regular vegetable intake, and �50% of students had no high-fatvegetable intake; therefore, it was necessary to combine some quartiles.bRemoved one outlier (�10 servings from all other values).M, mean; SD, standard deviation.

Am J Prev Med 2003;24(1) 59

recipe preparation in the virtual kitchen, the tailoringof goal setting to baseline dietary assessment, thetailoring of decision making to baseline reports of FJVpreferences and outcome expectancies, or other as-pects of Squire’s Quest! Further tests of this technologyshould systematically vary these components to eluci-date how this new technology influences mediatingvariables and behavioral outcomes.

Health educators can benefit from more contact withdevelopers who create games for children and byintegrating theoretical behavioral frameworks into ed-ucational games. Creating such educational games isvery expensive, requiring large teams of educational,dietary, and behavioral professionals, with subcontract-ing for professional artists and programmers. Periodi-cally updating the program, based on feedback fromchildren and teachers, could enhance the program butwould add to cost. Alternatively, the possible conver-sion of this educational technology to the Internetholds out the promise of reaching large numbers ofindividuals, thereby minimizing marginal cost per newparticipant. Some technologic challenges need to beovercome (e.g., the lengthy download time that dis-courages participation) before a smooth transitionfrom CD-ROM to the Internet can be made. Futureuses of Squire’s Quest! could include refinement andcontinued use in the classroom or conversion to anindividual sequential game that does not have definedsessions on CD-ROM or the Internet.

The reliability-related intraclass correlations across 4days of consumption were modest overall, but higherfor FJV combined. Reliability in dietary assessment is, inlarge part, a function of the number of days of assess-ment.44 It would be very challenging to collect �4 daysof data from large groups of children in school settingsusing computerized procedures. Other limitations ofthis study include the limited completion of all tensessions in the spring implementation (but suggestseven larger changes may have been attained if higherparticipation were achieved) and the dietary assessmentthat relied solely on self-report.

Squire’s Quest! demonstrates that PEMT games caninduce dietary behavior change among elementaryschool children. Further research is warranted to iden-tify key components in, and the duration of, effective-ness, as well as the generalizability of such games.

We are grateful to Colin McKay of SMILEX; Bruce Blausen ofBMC Software, and Steve Hite, Mukesh Taylor, and TomRobinson of Think Software, for their software programming;Electric Paintbrush and Sasha Fernandez, private entrepre-neur, for their artwork; Brenda Congden, for development ofcreative content; Linda Zelley, MS, RD, for development ofnutrition content; Felica Bradford, for data assessment; andKathy Watson MS, MPH, for data analysis.

This research was funded largely by the National Institutesof Health, grant R01 CA-75614. This work is also a publication

of the U.S. Department of Agriculture (USDA/ARS) Chil-dren’s Nutrition Research Center, at Department of Pediat-rics, Baylor College of Medicine, Houston, Texas, funded inpart by the USDA/ARS (Cooperative Agreement 58-6250-6001). The contents of this publication do not necessarilyreflect the views or policies of the USDA, nor does mention oftrade names or organizations imply endorsement by the U.S.government.

References1. Fraser GE, Shavlik DJ. Ten years of life: is it a matter of choice? Arch Intern

Med 2001;161:1645–52.2. Potter JD, ed. Food, nutrition and the prevention of cancer: a global

perspective. Washington DC: World Cancer Research Fund and AmericanInstitute for Cancer Research, 1997.

3. Joshipura KJ, Hu FB, Manson JE, et al. The effect of fruit and vegetableintake on risk for coronary heart disease. Ann Intern Med 2001;134:1106–14.

4. Tuomilehto J, Lindstrom J, Eriksson JG, et al. Prevention of type 2 diabetesmellitus by changes in lifestyle among subjects with impaired glucosetolerance. N Engl J Med 2001;344:1343–50.

5. Purba MB, Kouris-Blazos A, Wattanapenpaiboon N, et al. Skin wrinkling:can food make a difference? J Am Coll Nutr 2001;20:71–80.

6. Colditz G, Frazier AL. Models of breast cancer show that risk is set by eventsof early life: prevention efforts must shift focus. Cancer Epidemiol Biomar-kers Prev 1995;4:567–71.

7. Birch LL. Psychological influences on the childhood diet. J Nutr 1998;128(suppl 2):407–10.

8. Resnicow K, Smith M, Baranowski T, Baranowski J, Vaughan R, Davis M.Two-year tracking of children’s fruit and vegetable intake. J Am Diet Assoc1998;98:785–9.

9. Nicklas TA, Webber LS, Johnson CC, Srinivasan SR, Berenson GS. Foun-dations for health promotion with youth: a review of observations from theBogalusa Heart Study. J Health Educ 1995;26(suppl):18–26.

10. Kelder SD, Perry CL, Klepp KI, Lytle LL. Longitudinal tracking ofadolescent smoking, physical activity, and food choice behaviors. Am JPublic Health 1994;84:1121–6.

11. Baranowski T, Mendlein J, Resnicow K, Frank E, Cullen KW, Baranowski J.Physical Activity and Nutrition (PAN) in children and youth: An overviewof obesity prevention. Prev Med 2000;31(suppl):S1–S10.

12. Baranowski T, Lin LS, Wetter DW, Resnicow K, Hearn MD. Theory asmediating variables: Why aren’t community interventions working asdesired? Ann Epidemiol 1997;7:89–95.

13. Baranowski T, Anderson C, Carmack C. Mediating variable framework inphysical activity interventions. How are we doing? How might we do better?Am J Prev Med 1998;15:266–97.

14. Hearn M, Baranowski T, Baranowski J, et al. Environmental influences ondietary behavior among children: availability and accessibility of fruits andvegetables enable consumption. J Health Educ 1998;29:26–32.

15. Resnicow K, Hearn MD, Smith M, et al. Psychosocial correlates of fruit andvegetable consumption. Health Psychol 1997;16:272–6.

16. Kirby S, Baranowski T, Reynolds K, Taylor G, Binkley D. Children’s fruitand vegetable intake: socioeconomic, adult, child, regional, and urban-rural influences. J Nutr Educ 1995;27:261–71.

17. Reynolds KD, Franklin FA, Binkley D, et al. Increasing the fruit andvegetable consumption of 4th graders: results from the High 5 project. PrevMed 2000;30:309–19.

18. Luepker RV, Perry CL, McKinlay SM, et al. Outcomes of a field trial toimprove children’s dietary patterns and physical activity: the Child andAdolescent Trial for Cardiovascular Health (CATCH). The CATCH collab-orative group. JAMA 1996;275:768–76.

19. Baranowski T, Davis Hearn M, Resnicow K, et al. Gimme 5 fruit, juice andvegetables for fun and health: outcome evaluation. Health Educ Behav2000;27:96–111.

20. Resnicow K, Davis M, Smith M, et al. Results of the Teach Well WorksiteWellness Program. Am J Public Health 1998;88:250–7.

21. Davis M, Baranowski T, Resnicow K, et al. Gimme 5 fruit and vegetables forfun and health: process evaluation. Health Educ Behav 2000;27:167–76.

22. Morris JM, Owen GS, Fraser MD. Practical issues in multimedia userinterface design for computer-based instruction. In: Reisman S, ed. Multi-


media computing, preparing for the 21st century. Harrisburg, PA: IdeaGroup Publishing, 1996:65–105.

23. Shavinina LV. Interdisciplinary innovation: psychoeducational multi mediatechnologies. New Ideas Psychol 1998;16:189–204.

24. Paperny DM, Aono JY, Lehman RM, Hammar SL, Risser J. Computer-assisted detection and intervention in adolescent high-risk health behav-iors. J Pediatr 1990;116:456–62.

25. Bosworth K, Bustafson D, Hawkins RP, BARN Research Group. The BARNsystem: use and impact of adolescent health promotion via computers.Comput Human Behav 1994;10:467–82.

26. Prochaska JJ, Zabinski MF, Calfas KJ, Sallis JF, Patrick K. PACE�, interac-tive communication technology for behavior change in clinical settings.Am J Prev Med 2000;19:127–31.

27. Turnin MC, Bourgeois O, Cathelineau G, et al. Multicenter randomizedevaluation of a nutritional education software in obese patients. DiabetesMetab 2001;27:139–47.

28. Tate DF, Wing RR, Winett RA. Using Internet technology to deliver abehavioral weight loss program. JAMA 2001;285:1172–7.

29. Winzelberg AJ, Eppstein D, Eldredge KL, et al. Effectiveness of aninternet-based program for reducing risk factors for eating disorders. JConsult Clin Psychol 2000;68:346–50.

30. Anderson ES, Winett RA, Wojcik JR, Winett SG, Bowden T. A computerizedsocial cognitive intervention for nutrition behavior: direct and mediatedeffects on fat, fiber, fruits, and vegetables, self-efficacy, and outcomeexpectations among food shoppers. Ann Behav Med 2001;23:88–100.

31. Delichatsios HK, Friedman RH, Glanz K, et al. Randomized trial of a“talking computer” to improve adults’ eating habits. Am J Health Promot2001;15:215–24.

32. de Bourdeaudhuij I, Brug J. Tailoring dietary feedback to reduce fat intake:an intervention at the family level. Health Educ Res 2000;15:449–62.

33. Barker R, Gump P. Big school, small school. Stanford, CA: StanfordUniversity Press, 1964.

34. Currie CS, Elton RA, Todd J, Platt S. Indicators of socioeconomic status foradolescents: the WHO health behaviors in school aged children survey.Health Educ Res 1997;12:385–97.

35. Cullen K, Baranowski T, Baranowski J. Computer software design forchildren’s food intake recording: What do third, fourth, and fifth gradechildren like? J Nutr Educ 1998;30:405–9.

36. Baranowski T, Islam N, Baranowski J, et al. The Food Intake RecordingSoftware System (FIRSSt) is valid among 4th grade children. J Am DietetAssoc 2002;102:380–5.

37. Martin DC, Diehr PH, Perrin EB, Koepsell TD. The effect of matching onthe power of randomized community interventions studies. Stat Med1993;12:329–33.

38. Diehr P, Martin DC, Koepsell TD, Cheadle A. Breaking the matches in apaired t-test for community interventions when the number of pairs issmall. Stat Med 1995;14:1491–504.

39. Proschan MA. On the distribution of the unpaired t-statistic with paireddata. Stat Med 1996;15:1059–63.

40. Domel S, Baranowski T, Davis H, et al. Development and evaluation of aschool intervention to increase fruit and vegetable consumption among 4thand 5th grade students. J Nutr Educ 1993;25:345–9.

41. Hearn MD, Baranowski T, Baranowski J, et al. Environmental influences ondietary behavior among children: availability and accessibility of fruit andvegetables enable consumption. J Health Educ 1998;29:26–32.

42. Cullen KW, Zakeri I, Pryor EW, Baranowski T, Baranowski J. Goal setting isdifferentially related to change in fruit and vegetable consumption among4th grade children. Health Psychol In press.

43. Borra ST, Schwartz NE, Spain CG, Natchipolsky MM. Food, physicalactivity, and fun: inspiring America’s kids to more healthful lifestyles. J AmDiet Assoc 1995;95:816–23.

44. Baranowski T, Baranowski J, Doyle C, et al. Estimation of servings of fruitand vegetables and fat practices from adults’ seven day food records. J NutrEduc 1997;29:321–6.

Am J Prev Med 2003;24(1) 61

Documents

Squire’s Quest!