Normative Beliefs of School Children Concerning Regular Exercise zy Gaston Godin, Roy J. Shephard ABSTRACT The main purpose o f the study was to examine societal norms toward exercise as perceived by school children. To this end, the normative beliefs of specific referents (parents, teachers, and friends) and the motivation to comply with the beliefs o f these specific referents were measured in zyxwvutsrqpo 698 students o f both sexes attending grades seven-nine. Results indicated major changes in both normative beliefs and motivation to comply with “significant others” as the children move from grade seven to nine. There also were substantial differences between the results for active and sedentary children. Teachers seem to provide a Ftable norm. less influenced by the age. tex, and level of physical activity of the student. INTRODUCTION zyxwvuts Acquisition of a health habit is usually the result of both social and psychological factors. However, of the numerous external cues that influence behavior, none is more common or potent than the perception of “significant others.” During childhood, the parents have a major impact upon the develop- ment of precursors of adult social coping skills.’ With adoles- cence, the influence of the parents becomes powerfully supple- mented by the perceived standards of peers, teachers, and the mediaa2In the case of physical activity, the involvement of both children and youth appears related to parental encouragement of parti~ipation.~,~ Support received from the family, peer groups, teachers and coaches, plus the oppor- tunity to participate in organized sports from an early age create favorable conditions for continuation of an active life- style into the adult years’s6 Most educational efforts to date have lacked a sound theoretical framework, being based upon a simplistic knowl- edge-habit model.’ Thus, it is not surprising that general types of school health promotion programs have had only minimal effect on students’ health behavior.2*8-i0 Indeed, the natural channels of communication about health behavior have yet to be described adequately in the context of lifestyle modification programs.” Learning activities should be perceived as relevant, with personal meaning for the student. However, children differ in the degree to which their behaviors are guided by external “role-models” such as parents, friends, and teachers. It is therefore important to establish the per- ceived social norm for a given health behavior, and to measure the desire to comply with that norm.” This study examined societal norms toward exercise as perceived by school children. Emphasis was placed on the normative beliefs of specific referents such as parents, teachers, and friends and the motivation to comply with the beliefs of these specific referents. The resultant data have implications for pedagogy, particularly the adaptation of health (exercise) promotional tactics to the specific interests and values of young students. METHOD Subjects Participants were 698 junior high school students of both sexes attending public school grades seven-nine.* Physical education was a required subject in each of these grades. Two schools were selected for testing in collaboration with the roordinator of Physical and Health Education for the Borough of North York The initilil contart wa5 made by the coordinator and acces5 to student5 wa$ obtained through the cooperation of the Physical and Health Education departments within the two schools Data Collection Subjects completed a standard questionnaire during a regular physical and health education class. The quality of wording and vocabulary were tested in preliminary studies. The format of questions was adjusted after pilot trials of the proposed instrument at another school. Normative Beliefs and Motivation to Comply Individual questions were developed as suggested by Ajzen and Fishbein. I‘ This behavioral model assumes that an individual considers the normative expectations of various “others” and it also assesses the individual’s motivation to comply with such referents. In the definitive instrument, three normative beliefs were measured by having subjects agree or disagree with the proposition: “I think that doing active sports or vigorous physical activities a few times a week, during my leisure time, is something that my (parents, teachers, friends) believe I should do.” Motivation to comply was evaluated by three parallel questions of the type “I would like to do active sports or vigorous physical activities the way my (parents, teachers, friends) think I should do.” Responses to each question were rated on a five-point Likert scale (+ 2 to -2). The two week test-retest reliability coefficient” of the summed product of normative beliefs and motivation to comply was .67. Current Physical Activity Habits The physical activity habits of the sample were assessed using a questionnaire developed by Godin. l6 A weekly activity score was calculated for each child by summing indices based upon the reported weekly frequency of exercise at three inten- sity levels (strenuous, moderate, and mild) multiplied by the corresponding anticipated multiples of resting energy expendi- ture (nine, five, or three METs).* The two week test-retest reliability coefficient for this instrument was r = 0.84 and, in terms of validity, the scores reported by child and adolescent swimming competitors differed significantly from that reported by average children. ANALYSES Multivariate analyses of variance (MANOVA) compared normative beliefs and motivation to comply among children differing in exercise behavior, sex, and school grade, using Hotelling’s criteria.** One-way analyses of variance (ANOVA) were determined for each subcomponent of normative belief and motivation comply. RESULTS Activity Habits The children reported participating in strenuous exercise 3.9 3.0 times per week, in moderate exercise 4.4 * 2.7 times per week, and in light exercise 4.4 zyxw f 3.0 times per week. The overall physical activity score (Table 1) averaged 70.4 * 36.2. *MET means ratio to resting metabolic rate “For example the mean score vectors of active and sedentary children were tested to determine if a significant difference existed between their respective controls In analyses where a 5ignificant difference WAC ohsrrvrd simdtClneouF confidence interval5 were applied to detect wgniflcdnt differcmcw hetween active and sedentary children on individual items JOSH December 1984, Vol. 54, No. 11 443