Participants
A total of 171 students aged 13 to 16 years old and attending co-educational secondary schools voluntary participated in the present study. All secondary schools in Singapore were invited to participate in the study via e-mail through publicly available contact details. The research team also attended a meeting with school principals to invite their schools to participate in the study. The three secondary schools whose principal granted permission were included in the study, and approval for data collection at the schools was also obtained from Singapore Ministry of Education. After the department head and PE teachers selected two classes at their school, a researcher attended their PE lessons and explained the risks and benefits of the study to the students by distributing information sheets and informed consent forms. They also had the opportunity to ask questions about the study requirements prior to completing and signing an informed consent form. Students were asked to pass the information sheets and informed consent forms to their parents and discuss their participation in the study with them. Students who gave their consent and returned the informed consent forms signed by their parent were eligible to participate in the study. Parents who consented were asked to complete a brief demographics and socioeconomic status questionnaire. Students who did not return the signed informed consent forms or could not participate in PE lessons were excluded from the study. They participated in PE lessons together with their classmates, but did not do any tasks for the study.
The pilot study by Chatzisarantis et al. [
21] reported that their TPB-based intervention with mixed (between and within-subjects) design had a small effect on adolescents’ PA participation (ƞ
p2 = .01) and large effects on their beliefs (ƞ
p2 = .15). Therefore, a conservative small effect size of ƞ
p2 = .02 (
f = .14) was adopted to calculate the total sample size with G*Power version 3.1.9.2 [
22]. It was estimated that a sample of 174 participants was required to achieve a power of .90 at the alpha level of .05 for the repeated within-subjects analysis of 6 groups (2 classes at three schools) and 6 measurement points (3 measurement points in two conditions).
Procedure
Prior to the commencement of the study in each condition, students’ body compositions (e.g., height, body mass and body fat percentage) were measured at their school. Height was measured using a ruler attached to the wall to the nearest 0.5 cm, body mass and body fat percentage were measured with a body composition monitor (Omron Karada Scan: HBF-362). Students were also asked to complete a demographics sheet and a set of self-report measures of leisure-time PA and psychological variables (Pre-Baseline). After completing the questionnaire set, students were instructed to wear the GENEActiv (Original) accelerometer (Activinsights Ltd, Cambridgeshire, UK) on their non-dominant wrist for one week to objectively measure frequency and intensity of their PA level (Baseline). When students returned the accelerometer, they were also asked to complete another set of the same self-report measures (Baseline). Subsequently, students spent four weeks in the control or intervention condition. During the 4 weeks, an intervention program was conducted in PE lessons in the intervention condition. After the 4 weeks, students were requested to wear the accelerometer again for one week across two occasions (Post 1 and Post 2) with a 4 weeks interval period in between before measuring their body compositions and completing the same set of self-report measures. Students’ height, body mass and body fat percentage were measured at three occasions (Pre-Baseline, Post 1 and Post 2) in each condition (see Fig.
2).
Intervention and fidelity check
PE lessons were conducted twice a week over 4 weeks in both control and intervention conditions. In the control condition, PE teachers encouraged students to participate in PA during leisure time without delivering persuasive messages in PE lessons. An intervention program was implemented after the control period was over (see Fig.
2).
Prior to the intervention period, PE teachers were trained on how to deliver persuasive messages through a 3-h workshop. During the workshop, PE teachers were encouraged to express their concerns about delivering persuasive messages, researchers addressed those concerns to assist PE teachers to conduct the intervention at a sufficient level of proficiency. The persuasive messages targeted salient information related to the benefits and barriers associated with PA, which were identified based on the results of surveys conducted for students at Pre-Baseline in the control condition. Some examples of salient benefits were “health benefits”, “improving physical skills and fitness”, “fun and socialization” and “recommended physical activity level”, whereas salient barriers were “lack of time”, “lack of facilities or equipment”, “feeling tired” and “weather”.
In the intervention condition, PE teachers delivered the persuasive messages that targeted the salient benefits and barriers associated with PA at the last 5 to 10 min of each PE lesson. After PE teachers delivered the persuasive message, students were asked to answer a question about the message in each lesson (e.g., “Please explain briefly how do you know that MVPA improves your fitness levels?”). Researchers evaluated PE teachers’ fidelity of the interventions with a checklist (e.g., “The PE teacher provided clear instructions on this session’s message”) on a scale ranging from 1 (not at all) to 5 (very much).
Outcome measures
Students’ BMI were calculated based on their height and body mass, and overweight or obese students were identified based on Cole et al.’s classification [
23]. Students’ PA levels were measured objectively with GENEActiv (Original) accelerometers configured at 100 Hz, and subjectively with the International Physical Activity Questionnaire [
24] (e.g., “During the last 7 days, how many days did you do vigorous physical activities?”),and the Leisure-Time Physical Activity Participation Questionnaire [
25] (“How often did you do any regular activity long enough to sweat last week during your leisure time?”) on a 8-point scale ranging from 0 (
none) to 7 (
every day).
Students’ attitudes, intentions, subjective norms and perceived behavior control towards leisure-time PA were measured with a questionnaire based on the constructs from the TPB. Attitudes were assessed through five items with bipolar adjectives (enjoyable/unenjoyable, good/bad, useful/useless, interesting/boring, and beneficial/harmful) on a 7-point semantic differential scale [
26]. Behavioral intentions were assessed with three items (e.g., “For the next 4 weeks, I intent to do vigorous physical activity for at least 3 times [
at least 30 minutes] per week.”) based on a 7-point scale ranging from 1 (
unlikely) to 7 (
very likely) [
19]. Subjective norms were assessed through three items (e.g., “Most people who are important to me think I should do vigorous physical activity for at least 3 times [
at least 30 minutes] per week for the next 4 weeks.”) based on a 7-point scales ranging from 1 (
strongly disagree) to 7 (
strongly agree) [
21]. Perceptions of behavior control were measured using three items (e.g., “I feel in complete control over doing vigorous physical activities for at least 3 times [
at least 30 minutes] per week for the next 4 weeks.”) based on 7-point scale ranging from 1 (
completely false) to 7 (
completely true) [
19].
Participants were also asked to compare intention, determination and willingness to engage in another activity during their leisure time with vigorous PA and answer three questions (e.g., “How much do you intend to do the other activity for at least 30 minutes, 3 times a week for the next 4 weeks during your leisure time?”) based on a 7-point scale, ranging from 1 (not at all) to 7 (very much).
Goal conflict and facilitation were assessed through five items (e.g., “To what extent is the other activity likely to prevent you from doing vigorous physical activity for at least 30 minutes, 3 times a week for the next 4 weeks during your leisure time?”) based on a 7-point scale ranging from 1 (not at all) to 7 (very much).
Data analysis
To examine the first research hypothesis (
H
1
), a 2 (type of conditions: control vs. intervention) × 3 (measurement points: Baseline, Post 1 and Post 2) repeated ANOVA will be conducted on students’ MVPA levels. The second hypothesis (
H
2
) will be examined through a 2 (type of conditions: control vs. intervention) × 3 (measurement points: Pre-Baseline, Post 1 and Post 2) repeated ANOVA on students’ BMI levels. A path analysis will be conducted to examine the third hypothesis (
H
3
) with
Mplus (Version 8) [
27]. Student’s age, gender, socio-economic status, and prioritization of PA together with fidelity of intervention programs will be included as covariates in the analyses. Furthermore, correlation analyses will be conducted to examine the relationships between the subjective and objective measures of PA.
A cost effectiveness analysis of the program will also be conducted by focusing on the incremental cost-effectiveness of the intervention condition compared to the control condition. Incremental cost-effectiveness ratios will be calculated in terms of the incremental cost per unit BMI reduction. Differences in BMI between different measurement points (Pre-Baseline, Post 1 and Post 2) will also be calculated. Height will be added as a covariate in the analyses in order to account for the growth change in height over the data collection period. Incremental cost-effectiveness ratios will be calculated using standard techniques coupled with bootstrapping methodology and acceptability curves [
28]. Through these analyses, it will provide information about the probability that an intervention is cost-effective.