Reliability and Validity of a School-Based Physical... : Medicine & Science in Sports & Exercise (original) (raw)
Increasing the proportion of children and adolescents (youth) that are physically active remains a public health priority (24,25). Schools are increasingly under pressure to deal with physical inactivity because school-based prevention initiatives have the potential to reach a large number of youth at a critical developmental period (3). However, school stakeholders are not being provided with the tools or resources they need to make evidence-based programming decisions (3,18). The physical activity module of the School Health Action, Planning and Evaluation System (SHAPES) has been designed to provide school stakeholders with the evidence they need for guiding and evaluating school-based physical activity interventions.
SHAPES is a modular local data collection and feedback system designed for schools. Details of the data collection and feedback system concept have been described elsewhere (6). In brief, SHAPES was designed to (a) engage local health and education systems in planning, tailoring, and evaluating local population health initiatives based on evidence; (b) engage researchers in assessing environmental influences on behavior; and (c) provide a platform to study and facilitate effective knowledge exchange. Each SHAPES module consists of three components: 1) a low-cost, machine-readable questionnaire designed to collect data about a health behavior from all grade 6-12 students in a school; 2) a computer-generated feedback report with school-specific results; and 3) a school administrator questionnaire designed to collect data about school programs and policies related to the behavior.
SHAPES currently consists of two modules, the tobacco module and the physical activity module. A third module, which addresses eating behavior, is being developed. The first SHAPES module was designed to address youth smoking behavior; the reliability and validity of this module has already been established (9). Since 2000, over 185,000 students in more than 650 schools in Canada have completed the tobacco module in both researcher- and stakeholder-initiated projects (including national and provincial (state) government ministries and several local public health units). Recently, a module has been designed to address youth physical activity.
SHAPES PHYSICAL ACTIVITY MODULE QUESTIONNAIRE
SHAPES was designed to be used for multiple large-scale school-based data collections. Thus, there were numerous issues to consider in the development of the student questionnaire. A machine-readable format facilitates large-scale data collections by enabling fast and accurate processing of completed questionnaires while minimizing labor costs and transcription errors. To minimize both shipping and printing costs, it was preferable that the questionnaire be no more than four pages long. In addition, items were selected to balance administration time with providing sufficient detail to support planning and evaluation.
The physical activity questionnaire consists of 45 multiple-choice questions presented in a four-page machine-readable booklet. Two items request 7-d recall of vigorous physical activity (VPA) and moderate physical activity (MPA), respectively. VPA was defined as "jogging, team sports, fast dancing, jump-rope, and any other physical activities that increase your heart rate and make you breathe hard and sweat." MPA was defined as "lower intensity physical activities such as walking, biking to school, and recreational swimming." Responses are provided by indicating the number of hours (0-4 h) and 15-min increments (0-45 min) that each type of physical activity was performed for each day of the previous week. Thus, intensity, duration, and frequency data are collected, and weekday versus weekend analyses are possible. These physical activity items were designed to enable comparisons with results from previous Canadian national physical activity surveys (7) that classified physical activity levels based on energy expenditure (kcal·kg−1·d−1). In addition, the SHAPES items require less space than checklist-type measures and subsequently enable other items to be included in the questionnaire.
Items were included not only to measure self-reported physical activity levels but also to inform the planning and evaluation of school programs and policies. In addition to the two core physical activity items, items ask about participation in physical activities (e.g., physical education, strength training, intramural sports, varsity sports, commuting to school), sedentary activities (e.g., watching television, playing video games, homework), social influences (e.g., peer and parental influences), school environment (e.g., opportunities to participate in physical activity, indoor, and outdoor facilities), self-perceptions of weight status and athletic ability, height, weight, and demographics (e.g., age, grade, sex). The height and weight items were worded identically to those used in the Youth Risk Behavior Survey (5). However, in the Canadian context, both metric and imperial multiple-choice response options must be provided. Further, the items included a blank line prefaced by "My weight is" and "My height is," which participants could complete. Because of the different format of the response options, it was important to establish the test-retest reliability and criterion validity of the height and weight items in the SHAPES questionnaire.
Feedback from schools and stakeholders that have used the SHAPES tobacco module and expressed interest in using the physical activity module have indicated that items such as school connectedness and smoking behavior would provide information that is of interest to them. Although these topics may not generally be of interest for physical activity research, these items increase the potential usefulness of the questionnaire to schools and stakeholders. Thus, the physical activity questionnaire includes the core smoking behavior items. Similarly, the tobacco module questionnaire has been modified to include the core physical activity, height, and weight items, and both questionnaires include the school connectedness items. By administering each questionnaire to half of the students in a school, complete data for the core measures can be collected from the entire student population while simultaneously collecting more detailed data for each health behavior from half of the student population. This strategy would maximize the amount of data collected without increasing the burden on the schools. Increasing the value of the information provided by participating in SHAPES surveys increases the likelihood that school boards and schools will consent to participate in school-based data collections.
Pilot testing of the SHAPES physical activity module questionnaire included quantitative and qualitative studies to determine readability and comprehension. A copy of the complete questionnaire, as well as a sample feedback report, can be obtained online at www.shapes.uwaterloo.ca or by contacting the corresponding author. Establishing the psychometric properties of the SHAPES physical activity module student questionnaire is important for all potential users of SHAPES data, including researchers and education stakeholders. The present paper details the reliability and validity testing of the SHAPES physical activity questionnaire.
STUDY 1: TEST-RETEST RELIABILITY
The purpose was to assess the 1-wk test-retest reliability of the SHAPES physical activity questionnaire.
Methods: study 1
The test-retest study for the SHAPES physical activity questionnaire used convenience sample data from 2812 students (grades 9-12) from 15 Manitoba schools, 13 of which were selected because they were already participating in a multiprovince smoking survey affiliated with the members of the research team. Students completed the physical activity questionnaire in class time on two occasions in spring 2004. At time 1, teachers administered the questionnaires in classrooms using a common protocol that included standardized instructions to the students (N = 2812; within the 12 schools for which we had enrollment data for, this represented 24.3% of the total student population). One week later (time 2), the questionnaires were readministered to the same classes (N = 2518; 89.5% of time 1 participants). Students created a self-generated code to permit matching of time 1 and time 2 questionnaires.
The study employed active information with passive consent procedures to reduce demands on schools and to increase student participation rates. That is, researchers informed the parents of the students about the study via mailed letter, and asked them to call the public health contact person for their child's school (toll-free number accessible 24 h·d−1) if they declined participation. The University of Waterloo and University of Manitoba offices of research ethics and appropriate school board and public health ethics committees approved all procedures.
Statistical analysis: study 1.
Data from 1636 students (58.2% of the time 1 students) were included in the statistical analyses. Data from 294 students were excluded because they did not complete both the time 1 and time 2 questionnaires, and 882 students were excluded because of an inability to link their time 1 and time 2 data. There was a significant gender difference in the students with matched versus unmatched data (54% of the unmatched cases were male compared with 45% for the matched cases, P < 0.05). Unmatched cases also tended to be slightly older (16.3 ± 1.5 vs 16.0 ± 1.6 yr). These two factors may account for the significantly greater height and weight for the unmatched versus matched cases. However, there were no significant differences in self-reported time spent performing moderate- or vigorous-intensity physical activity (P = 0.24 and 0.08, respectively).
Variables used to calculate derived variables were excluded from the analyses, as were questionnaire items with responses that were not expected to be similar from 1 wk to the next or response formats that requested the participant to choose all responses that applied. Body mass index (BMI, kg·m−2) was calculated as weight (kg) divided by height (m) squared. Weekly time spent performing VPA and MPA were calculated by summing the responses from the two 7-d recalls that reported time spent performing VPA and MPA, respectively, each day for the last 7 d. Weekly time spent performing moderate to vigorous physical activity (MVPA) was calculated by summing the weekly time spent performing VPA and MPA. Energy expenditure (KKD, kcal·kg−1·d−1) to classify physical activity levels was estimated from the two 7-d recalls, assuming MPA = 4 METs and VPA = 6 METs. Respondents' physical activity levels were classified as inactive (< 3 KKD), moderately active (3-7.9KKD), and active (8 KKD). Weekly "screen time" was derived by summing the responses from the 7-d recall of time spent watching TV, movies, playing video/computer games, surfing the Internet, or talking on the phone.
Test-retest reliability was assessed for agreement beyond what would be expected by chance alone by using the kappa statistic for categorical responses and weighted kappa statistic for ordinal responses. Interpretation of values of kappa and weighted kappa were based on the classification system developed by Landis and Koch (13), where ≤ 0.09 indicated poor agreement, 0.10-0.20 indicated slight agreement, 0.21-0.40 indicated fair agreement, 0.41-0.60 indicated moderate agreement, 0.61-0.80 indicated substantial agreement, and 0.81-1.00 indicated almost perfect agreement. ANOVA was used to determine whether there were significant differences in the test-retest reliability of the questionnaire by gender and by grade, respectively. All statistical analyses were performed using SAS (version 9.1; SAS Institute Inc., Cary, NC).
Results: study 1
Table 1 presents the means and standard deviations of the physical activity variables, height, weight, and body mass index by gender and combined. Table 2 summarizes the kappa and weighted kappa coefficients for 1-wk test-retest reliability by item domain. Kappa and weighted kappa coefficients ranged from 0.30 to 0.99. Overall, the mean kappa/weighted kappa coefficient for the 49 items was 0.57 ± 0.24. The coefficients for 8% of the items were ≥ 0.81, which indicated almost perfect agreement, whereas the coefficients for 33% of the items were ≥ 0.61, which indicated substantial agreement. Another 41% of items had coefficients in the moderate agreement range (0.41-0.60), and the remaining 18% of items were in the fair agreement range (0.21-0.40). Table 3 summarizes the weighted kappa coefficients for the variables derived from SHAPES physical activity questionnaire responses. There was fair agreement for VPA, MPA, and MVPA; however, classification into physical activity levels resulted in moderate agreement. Agreement for weekly screen time and BMI were moderate and substantial, respectively. The mean kappa/weighted kappa coefficient for the 49 items and the derived variables was 0.58 ± 0.17. Results of the ANOVA analyses showed borderline significant effects for gender (P = 0.07) and grade (P = 0.05). The mean and standard deviation of the kappa/weighted kappa coefficients by gender and grade are shown in Table 4.
Means and standard deviations of height, weight, body mass index, and physical activity of study 1 participants based on self-report, presented by gender and combined.
Kappa and weighted kappa coefficients for 1-wk test-retest reliability of items from the SHAPES physical activity questionnaire.
Weighted kappa coefficients for 1-wk test-retest reliability of derived variables from the SHAPES physical activity questionnaire.
Kappa/weighted kappa coefficients of the SHAPES physical activity questionnaire items and derived variables, by gender, by grade, and combined.
Discussion: study 1
The majority of items on the SHAPES physical activity questionnaire had acceptable reliability. The results suggest that the test-retest reliability of the physical activity, height, and weight items from the SHAPES questionnaire were comparable with other questionnaires for youth. For instance, the 2-wk test-retest reliability of the Adolescent Physical Activity Recall Questionnaire (APARQ) and the World Health Organization Health Behavior in School-Aged Children (WHO HBSC) survey were assessed in 226 grade 8 and 10 students (1,2). The APARQ summarized physical activity into a three-category measure of activity, and weighted kappa coefficients ranged from 0.33 to 0.71, depending on grade and gender (2). The WHO HBSC assessed frequency and duration of VPA performed outside school hours (1). Weighted kappa coefficients ranged from 0.22 to 0.60, depending on grade and gender. The 1999 Youth Risk Behavior Survey included five physical activity questions, as well as height and weight (4,5). Kappa coefficients for 2-wk test-retest reliability among their 4619 students in grades 9-12 ranged from 46.7 to 84.8 for the physical activity questions (4), whereas the kappa coefficient for a three-level BMI classification based on self-reported height and weight was 0.84 (5).
Lower test-retest reliability estimates indicated that there was a greater proportion of responses to the same item that did not match between times 1 and 2. However, differences in response to an item may reflect real change rather than response error when calculating kappa and weighted kappa coefficients. For example, a student could report at time 1 that he/she did not perform any moderate-intensity physical activity in the last 7 d and then report at time 2 that he/she performed 120 min of moderate physical activity in the last 7 d. Such responses would be inconsistent yet accurate if the student did indeed perform 120 min of moderate-intensity physical activity during the test-retest interval and not before. Week-to-week variation in the amount and/or intensity of physical activities performed, as well as time spent in sedentary activities, would be expected to some extent. Similarly, items in the school environment and school connectedness categories asked for subjective responses, such as opinions (e.g., the degree to which they agreed or disagreed that the indoor physical activity facilities at school met their needs) and feelings (e.g., the degree to which they agreed or disagreed that they felt close to people at school), which may also vary substantially from week to week. However, the analyses assumed that such inconsistencies were due to response error. Thus, kappa and weighted kappa coefficients may have been conservative estimates for some items.
A limitation of this study was that it was conducted among grade 9-12 students, even though the questionnaire was designed to be potentially used by students in grades 6-12. Additional research is required to determine the test-retest reliability of the questionnaire among students in grades 6-8.
STUDY 2: CRITERION VALIDITY STUDY
The purpose was to assess the criterion validity of two 7-d physical activity recalls and height and weight items of the SHAPES physical activity questionnaire using accelerometers and measured height and weight.
Methods: study 2
The criterion validity study for the SHAPES physical activity questionnaire used data from 67 students (grades 6-8, 10, and 12) collected in March-June 2005 from one Ontario private school. Manufacturing Technology Inc. (MTI) Actigraph model AM7164 (formerly the Computer Science and Applications, Inc.; CSA 7164) accelerometers were used to objectively assess physical activity. The MTI accelerometer is a small, lightweight uniaxial accelerometer designed to detect vertical acceleration ranging in magnitude from 0.05 to 2.00 g, with a frequency response of 0.25-2.50 Hz (i.e., movement outside of normal human motion is filtered out electronically). The filtered acceleration signal is digitized and the magnitude is summed over a user-specified time interval. At the end of each interval, the summed value or activity "count" is stored in memory and the integrator is reset. The current study used a 1-min time interval. The MTI/CSA accelerometer has previously demonstrated adequate reliability and validity for assessing physical activity (10,16,22).
In the present study, accelerometers were distributed during kinesiology class for the grade 12 students and during physical education class for the remaining students, with one wave of data collection for each grade (the duration of each wave of data collection varied from 7 to 9 d to accommodate teachers' schedules and lesson planning). On the first day of data collection, students in the participating class were provided with accelerometers that were initialized to begin recording data at a specified date and time. Students were instructed to wear the accelerometers during waking hours, except during water activities such as bathing and swimming, for each day of data collection. Consistent with previous studies, accelerometers were attached to adjustable elastic belts and worn over the right hip (23). Written and verbal instructions were provided and the elastic belts were adjusted for each participant. At the time of distribution, participants were given a daily logbook to record the time the accelerometer was worn and to provide information about any physical activity that was performed while the accelerometer was not worn (e.g., swimming). On the last day of data collection for each wave, the accelerometers were collected and SHAPES physical activity questionnaires were administered. Students were then asked to remove their footwear, and height and weight were measured using a standard protocol (5). Data from the accelerometers were then downloaded and saved as electronic files for subsequent data reduction and analysis. One parent or guardian provided informed written consent, and assent was obtained from each participant. Draws for gift certificates to a local shopping center were used as an incentive for participants to return the accelerometers and encourage compliance. Participants received a copy of their accelerometry results. The study was approved by the ethics review board at the University of Waterloo and by the school principal.
Statistical analysis: study 2.
To be included in the statistical analyses, participants must have worn the accelerometer for at least 10 h·d−1 for a minimum of 5 d. This was consistent with Trost et al. (23), who showed that at least 3 d of measurement in grade 1-6 students would be required to achieve a reliability of 0.70, whereas 5 d of measurement would be required in grade 7-12 students. Consistent with previous studies (17), 10 min or more of consecutive zero-activity counts was considered time that the accelerometer was not worn. Of the 67 participants, 53 (28 boys, 25 girls) had usable accelerometer data and completed the questionnaire. Although all participants had their height and weight measured, only 49 (23 boys, 26 girls) responded to both height and weight items on the questionnaire.
Accelerometer variables used in the statistical analyses were calculated using previously established cut points and equations (16). MPA and VPA were defined using the cut points of Puyau et al. (16), with MPA between 3200 and 8199 counts·min−1 and VPA ≥ 8200 counts·min−1. The minutes of MPA performed for complete days (at least 10h·d−1 for at least 5 d) were summed and divided by the number of complete days to calculate the average daily time spent performing MPA. The average daily time spent performing VPA and MVPA, respectively, were calculated similarly. Estimates of energy expended (KKD, kcal·kg−1·d−1) from MVPA were calculated using the equation derived by Puyau et al. (16): activity energy expenditure (kcal·kg−1·min−1) = 0.0183 + 0.000010 (counts).
Self-reported average daily time spent performing MPA and VPA were calculated by summing the responses from each day from the respective questionnaire item and dividing by seven. The weekly MPA and VPA performed were summed and divided by seven to calculate the average daily time spent performing MVPA. Self-reported energy expenditure from MVPA was calculated from the time spent performing MPA and VPA, and assuming that MPA = 4 METs and VPA = 6 METs. BMI was calculated as weight (kg) divided by height (m) squared for both self-reported and measured BMI (kg·m−2). Reference data from the CDC growth charts were used to classify students as "at risk for overweight" (BMI > 85th percentile, but < 95th percentile) and "overweight" (BMI > 95th percentile) (12).
The relationship between self-reported and objectively measured variables was determined using Spearman rank-order correlations. To assess the accuracy of self-report, the measured values were subtracted from the SHAPES values to obtain the difference in values. All statistical analyses were performed using SAS (version 9.1; SAS Institute Inc., Cary, NC).
Results: study 2
Table 5 presents the means and standard deviations of the physical activity variables, height, weight, and BMI by gender and combined. The correlation between self-reported and accelerometer-measured time spent performing MVPA was modest (r = 0.44) but significant (P < 0.01). Although the correlation for time spent performing MPA was significant, it was not for time spent performing VPA (see Table 6). Mean differences for time spent performing MPA, VPA, and MVPA were positive, indicating that on average self-reported values were higher than measured values (see Table 7). Self-reported time spent performing physical activity was higher compared with measured values for a large majority of the participants (see Table 7).
Means and standard deviations of height, weight, body mass index, and physical activity of study 2 participants based on self-report, presented by gender and combined.
Spearman correlations between self-reported and objectively measured variables.
Comparison between self-reported (S) and objectively measured (M) variables.
Correlations between self-reported and measured height, weight, and BMI, respectively, were high and significant (see Table 6). Although the difference in self-reported and measured height was >10 cm for two participants, the difference for the remainder of the participants ranged from −5.3 to 2.9 cm. Mean differences for height, weight, and BMI were negative, indicating that, on average, self-reported values were lower than measured values (see Table 7). However, self-reported height, weight, and BMI were not consistently higher or lower than measured values (see Table 7). Based on BMI from self-reported height and weight, 18% of students were classified as at risk for overweight, and 4% were classified as overweight, whereas 22% were at risk for overweight, and 4% were overweight based on measured values.
Discussion: study 2
The results indicated that responses to the core SHAPES physical activity items correlated with objectively measured physical activity. There was a significant correlation between SHAPES-estimated and accelerometer-estimated time spent performing MPA, MVPA, and energy expenditure; however, the correlation for time spent performing VPA was not significant. Although the significant correlation coefficients were modest, they were comparable with the results of previous studies that used accelerometers to validate self-report questionnaires in children and adolescents (8,11). For instance, the Physical Activity Questionnaire for Older Children (PAQ-C) and Physical Activity Questionnaire for Adolescents (PAQ-A) were moderately correlated with a Caltrac accelerometer in children (r = 0.39, P < 0.05) (8) and adolescents (r = 0.33, P < 0.05) (11), respectively. Similarly, a modified weekly activity checklist correlated with an MTI accelerometer (r = 0.30, P < 0.01) in 109 participants aged 8-16 yr (15). However, not all studies found significant correlations. Using the Children's Leisure Activities Study Survey (CLASS) 7-d checklist for physical activity recall, Spearman correlations for self-reported and accelerometer-measured time spent in moderate, vigorous, and total physical activity were all nonsignificant (21). Generally, it appears that the strength of correlations between the SHAPES physical activity questionnaire and an accelerometer were as robust as other youth 7-d physical activity recalls that were significantly correlated with accelerometry.
Although the results indicated that a large proportion of students had greatly overreported the amount of time spent performing physical activity, previous studies have observed similar findings. For instance, 320 students in grades 6-8 reported approximately 146 min·d−1 of MVPA on a 3-d physical activity recall compared with approximately 28min·d−1 as determined by MTI accelerometers (14). By comparison, students in the current study performed approximately 129 min·d−1 of MVPA according to self-report and approximately 31 min·d−1 according to MTI accelerometers. The degree of overreporting suggests that the SHAPES questionnaire should not be used to classify student physical activity levels based on time or energy expenditure. However, the significant correlation for MVPA and the tendency for the vast majority students to overreport suggest that it would be a suitable outcome measure for comparing groups.
Although accelerometers are frequently used as criterion measures of physical activity in studies to validate self-report questionnaires, there are a number of limitations associated with accelerometry. They have a limited ability to assess bicycling, locomotion on a gradient, or other activities with limited torso movement, and they may not be sensitive to many of the complex movement patterns exhibited by children during free play (10,22). In addition, the use of accelerometer count cut points to determine MPA and VPA may create measurement error. Similarly, converting accelerometer counts to units of energy expenditure may result in additional measurement error. Nevertheless, accelerometers provide an objective and nonreactive tool for assessing physical activity and are considered an acceptable means to validate self-report questionnaires (19).
The results of this study indicated that there was a strong correlation between SHAPES-estimated and measured height, weight, and BMI, respectively, and are similar to the results of previous studies. In a study of the 1999 Youth Risk Behavior Survey (YRBS), the Pearson correlation between self-reported and measured height, weight, and BMI for 4619 students in grades 9-12 ranged from 0.82 to 0.94, depending on grade and gender (5). The results of the National Health and Nutritional Examination Survey, Cycle III (NHANES III), which used an interview format, also showed similar results. The correlation between self-reported and measured weight ranged from 0.87 to 0.94, depending on gender or race, whereas the correlation for height ranged from 0.82 to 0.91 (20).
In this study, students underreported height and weight by an average of 1.2 cm and 0.3 kg, respectively. BMI values were an average of 0.3 kg·m−2 lower based on self-reported values compared with measured values, resulting in a similar percentage of students being classified as at risk for overweight and overweight. Results were similar for NHANES III. On average, self-reported height was 1.0 cm lower than measured height, and self-reported weight was 0.4 kg lower than measured weight (20). The use of self-reported height and weight resulted in the correct classification of weight status for 94% of adolescents. However, for the 1999 YRBS, students overreported their height by an average of approximately 6.9 cm and underreported their weight by an average of approximately 1.6 kg (5). Resulting BMI values were an average of 2.6 kg·m−2 lower when based on self-reported values compared with measured values. These findings suggest that the SHAPES questionnaire was comparable with, or slightly better than, other self-report measures of height and weight, and that the resulting BMI values would be suitable for classifying weight status.
This study had several limitations. Because of its informed consent procedure, students were aware that their height and weight would be measured, which may have led to more accurate reporting than in a typical SHAPES questionnaire administration. Further, missing data from students' absence or refusal to participate might have biased the results. However, without additional data, the direction of this bias cannot be predicted. Lastly, because of the small sample size, it was not possible to determine whether the results varied by grade and gender.
GENERAL DISCUSSION
Together, the results of these studies suggest that for large-scale school-based data collections, the SHAPES physical activity questionnaire had acceptable test-retest reliability among grade 9-12 students and acceptable validity for the core physical activity, height, and weight items among grade 6-12 students. Moreover, these results were comparable in reliability and validity with similar self-report measures for this age group (1,2,8,11). Thus, the SHAPES questionnaire appears suitable for assessing group-level physical activity and BMI of students in grades 6-12. Although some measures may not be suitable for assessing these variables at the individual level, this was not the purpose for which it was designed.
There were a number of items on the SHAPES questionnaire that provided information on physical activity behaviors and context related to physical activity whose validity has not been evaluated. For instance, it was not possible to validate the items pertaining to participation in intramural and interscholastic sports using accelerometry. However, it has been recognized that it is important not only to gather data on the amount of physical activity in which children and adolescent engage but also to gain an understanding of physical activity patterns and influences (21). Behavioral and contextual information is critical for the development of effective strategies to promote physical activity among children and adolescents (21). The importance of this type of information, the adequate test-retest reliability, and evidence from readability and comprehension studies support the inclusion of these items on the SHAPES questionnaire.
Changes to questionnaire items and additional reliability and validity testing will be considered based on the results of these studies and others, as well as feedback from schools, public health practitioners, and other stakeholders. This latter feedback is critical if SHAPES is used to provide feedback that guides planning and evaluation at the school level, as is intended. However, these results suggest that the current SHAPES physical activity questionnaire has adequate reliability and validity in a machine-readable format, which is a critical component to developing an effective local data collection and feedback system to facilitate the planning and evaluation of school-based interventions.
The Canadian Institutes of Health Research provided funding for study 1. The Canadian Cancer Society and National Cancer Institute of Canada's Centre for Behavioural Research and Program Evaluation supported both studies and funded study 2. The Lyle Hallman Health Promotion Institute helped purchase the accelerometers. S. L. Wong was supported by a Canadian Institutes of Health Research Canada Graduate Scholarship Doctoral Award.
We would like to acknowledge Roy Cameron, Tammy Cumming, Yolanda Dorrington, Keerat Grewal, Dexter Harvey, Mari Alice Jolin, Vanessa Rampersad, Lindsay Shantz, and Matthew VanderMeer for their contributions to these studies. We would also like to thank the school boards, schools, and students for their participation.
REFERENCES
1. Booth, M. L., A. D. Okely, T. Chey, and A. Bauman. The reliability and validity of the physical activity questions in the WHO health behaviour in schoolchildren (HBSC) survey: A population study. Br. J. Sports Med. 35:263-267, 2001.
2. Booth, M. L., A. D. Okely, T. N. Chey, and A. Bauman. The reliability and validity of the Adolescent Physical Activity Recall Questionnaire. Med. Sci. Sports Exerc. 34:1986-1995, 2002.
3. Botvin, G. J. Advancing prevention science and practice: Challenges, critical issues, and future directions. Prev. Sci. 5:69-72, 2004.
4. Brener, N., L. Kann, T. McManus, S. Kitchen, E. Sundberg, and J. Ross. Reliability of the 1999 Youth Risk Behavior Survey Questionnaire. J. Adol. Health 31:336-342, 2002.
5. Brener, N. D., T. Mcmanus, D. A. Galuska, R. Lowry, and H.Wechsler. Reliability and validity of self-reported height and weight among high school students. J. Adolesc. Health 32:281-287, 2003.
6. Cameron R., S. Manske K. S., Brown M.,Jolin D.,Murnaghan and C. Lovato. Integrating public health policy, practice, evaluation, surveillance, and research using local data collection and feedback systems: a prototype focused on schools. Am. J. Public Health (in press).
7. Craig, C. L., C. Cameron, and S. J. Russell. In: Beaulieu. Increasing physical activity: Supporting children's participation. Ottawa, Canada: Canadian Fitness and Lifestyle Research Institute, 2001, pp. 121-124.
8. Crocker, P. R., D. A. Bailey, R. A. Faulkner, K. C. Kowalski, and R. McGrath. Measuring general levels of physical activity: Preliminary evidence for the Physical Activity Questionnaire for Older Children. Med. Sci. Sports Exerc. 29:1344-1349, 1997.
9. Ekelund, U., M. Sjostrom, A. Yngve, et al. Physical activity assessed by activity monitor and doubly labeled water in children. Med. Sci. Sports Exerc. 33:275-281, 2001.
10. Freedson, P. S. Electronic motion sensors and heart rate measures of physical activity in children. J. School Health 61:220-223, 1991.
11. Kowalski, K. C., P. R. E. Crocker, and N. P. Kowalski. Convergent validity of the Physical Activity Questionnaire for Adolescents. Ped. Exerc. Sci. 9:342-352, 1997.
12. Kuczmarski, R. J., C. L. Ogden, L. M. Grummer-Strawn, et al. CDC Growth Charts: United States. Advance Data from Vital and Health Statistics; No. 314. (DHHS Pub. No. PHS 2000-12500-0431), Hyattsville, MD: National Center for Health Statistics, 2000.
13. Landis, J. R., and G. G. Koch. The measurement of observer agreement for categorical data. Biometrics 33:159-174, 1977.
14. McMurray, R. G., K. B. Ring, M. S. Treuth, et al. Comparison of two approaches to structured physical activity surveys for adolescents. Med. Sci. Sports Exerc. 36:2135-2143, 2004.
15. Mota, J., P. Santos, S. Guerra, J. C. Ribeiro, J. A. Duarte, and J. F. Sallis. Validation of a physical activity self-report questionnaire in a Portugese pediatric population. Ped. Exerc. Sci. 14:269-276, 2002.
16. Puyau, M. R., A. L. Adolph, F. A. Vohra, and N. F. Butte. Validation and calibration of physical activity monitors in children. Obes. Res. 10:150-157, 2002.
17. Riddoch, C. J., L. B. Anderson, and N. Wedderkopp. Physical activity levels and patterns of 9- and 15-yr-old European children. Med. Sci. Sports Exerc. 36:86-92, 2004.
18. Ringwalt, C., S. T. Ennett, A. A. Vincus, L. A. Rohrbach, and A. Simons-Rudolph. Who's calling the shots? Decision-makers and the adoption of effective school-based substance use prevention curricula. J. Drug Educ. 34:19-31, 2004.
19. Sirard, J. R., and R. R. Pate. Physical activity assessment in children and adolescents. Sports Med. 31:439-454, 2001.
20. Strauss, R. S. Comparison of measured and self-reported weight and height in a cross-sectional sample of young adolescents. Int. J. Obes. Relat. Metab. Disord. 23:904-908, 1999.
21. Telford, A., J. Salmon, D. Jolley, and D. Crawford. Reliability and validity of physical activity questionnaires for children: The Children's Leisure Activities Study Survey (CLASS). Ped. Exerc. Sci. 16:64-78, 2004.
22. Trost, S. G., D. S. Ward, S. M. Moorehead, P. D. Watson, and W. Riner. Validity of the computer science and applications (CSA) activity monitor in children. Med. Sci. Sports Exerc. 30:629-633, 1998.
23. Trost, S. G., R. R. Pate, P. S. Freedson, J. F. Sallis, and W. C. Taylor. Using objective physical activity measures with youth: How many days of monitoring are needed? Med. Sci. Sports Exerc. 32:426-431, 2000.
24. U.S. Department of Health and Human Services. Physical Activity and Health: A Report of the Surgeon General (USDHHS Pub. No. S/N 017-023-00196-5). Atlanta, GA: Centers for Disease Control and Prevention, 1996, pp. 4-6.
25. U.S. Department of Health and Human Services. Healthy People 2010: Understanding and Improving Health. 2nd Ed. (Stock No. 017-001-001-00-550-9). Washington, DC: U.S. Government Printing Office, 2000, pp. 17, 26-27.
Keywords:
CHILDREN; ADOLESCENTS; MEASUREMENT; SELF-REPORT; ACCELEROMETER; SURVEY
©2006The American College of Sports Medicine