Reliability and Validity of the SE-HEPA

Abstract

Objective. The purpose of this study was to examine the psychometric properties of the self-efficacy for healthy eating and physical activity measure (SE-HEPA) for preadolescents. Method. The reliability of the measure was examined to determine if the internal consistency of the measure was adequate (i.e., αs > .70). Next, in an effort to determine if a two-factor model was a better fit than a one-factor model, as hypothesized, an exploratory factor analysis in a confirmatory factor analysis framework was used to determine model fit. Additionally, the criterion-related validity of the measure was evaluated by conducting correlational analyses to determine if SE-HEPA scores were associated with preadolescent body mass index. Results. Consistent with the hypotheses, Cronbach’s alphas indicated good reliability for the measure (i.e., αs > .70) and factor analyses indicated the two-factor model provided a better fit than the one-factor model. Additionally, correlations revealed a significant relationship between the two factors (i.e., healthy eating, physical activity) and preadolescent body mass index. Conclusions. The SE-HEPA will allow researchers and clinicians to better understand self-efficacy for behaviors important to weight loss among preadolescents. This could, in turn, inform future efforts aimed at increasing behaviors that promote healthy weight status among this population within the context of family-based weight loss programs.

Keywords

health behavior obesity preadolescent self-efficacy weight management

In the past 20 years, pediatric obesity rates in the United States have increased sharply, leading to a dramatic increase in pediatric health problems (Ogden, Flegal, Carroll, & Johnson, 2002). For example, between 1980 and 2006, the number of obese children and adolescents (i.e., ages 2-19 years) increased from 6% to 16% (Ogden et al., 2002; Ogden, Carroll, & Flegal, 2003, 2008; Ogden, Yanovski, Carroll, & Flegal, 2007). Although there is no single cause of the current pediatric obesity epidemic, some have posited that the dramatic increase in overweight and obesity rates is partially because of the multifaceted nature of obesity (e.g., physical activity and eating patterns; Davidson, Marshall, & Birch, 2006; Raynor & Maier, 2006).

Fortunately, research indicates that multifaceted programs, including physical activity education, nutrition education, behavioral principles, and family involvement can prevent, and even ameliorate, many of the negative consequences of pediatric obesity and overweight (Jelalian, Markov-Wember, Bungeroth, & Burmaher, 2007). As one might imagine, these treatment strategies are most effective when the participating child and/or family members actually engage in the behaviors recommended during the treatment process (Steele, Steele, & Hunter, 2009). Consequently, research efforts have begun to focus on identifying ways to improve the likelihood that children and families will actually engage in behaviors important to weight loss. In fact, an expert committee for the assessment and treatment of pediatric obesity recently recommended that pediatric weigh loss programs tailor treatment efforts to fit each participating child’s level of “readiness to change” as a way to improve adherence among this population (Spear et al., 2007).

Several theories have been developed to better understand the concept of readiness to change, such as social cognitive theory (Bandura, 1977), transtheoretical model (Prochaska & DiClemente, 1983), and attribution theory (Kelly, 1967). These different frameworks can be perhaps most beneficial for understanding behavioral change through the examination of constructs that are amenable to change. For example, self-efficacy was adapted from Bandura’s (1977) social cognitive theory and can be defined as the confidence people have in their ability to change a particular behavior. Because self-efficacy appears to be amenable to intervention efforts, it has recently received a great deal of attention in the literature (e.g., Burke, Arkowitz, & Menchola, 2003). Numerous studies have examined self-efficacy for behaviors important to weight status in adult populations (e.g., Dallow & Anderson, 2003; Linde, Rothman, Baldwin, & Jeffery, 2006; Wiltink et al., 2007). Unfortunately, the lack of both condition-specific (i.e., weight loss) and child-specific measures of self-efficacy has limited the progress of this line of research. Although measures do exist, that are designed either to be child specific or condition specific (e.g., Motl et al., 2000), currently no single measure has been designed to capture both physical activity– and healthy eating–specific self-efficacy in children and adolescents.

A lack of condition- and child-specific measures of self-efficacy can be seen as especially problematic given that measures for this construct should include multiple domains, should be relevant to the situations and demands for the population of interest and that “one measure fits all” approaches are inappropriate (Bandura, 2005). This, in conjunction with the fact that preadolescence is a crucial time for developing dietary and physical activity habits (Kelder, Perry, Klepp, & Lytle, 1994), and these behaviors are related to weight status (e.g., Alinia, Hels, Tetens, 2009; Biddle, Gorely, & Stensel, 2004), indicates a need for a single child-specific measure of self-efficacy for weight status–relevant behaviors. Thus, the goal of the current investigation was to evaluate the psychometric properties of a self-report measure designed to assess self-efficacy for healthy eating and physical activity among a sample of preadolescent (i.e., the SE-HEPA).

The SE-HEPA measure was developed by Steele, Bindler, Power, and Daratha (2008) and is based on Motl et al.’s (2000) unidimensional measure of self-efficacy for exercise. The SE-HEPA is based on Bandura’s (1977) social cognitive theory (Motl et al., 2000) and was designed to evaluate a child’s/adolescent’s confidence in his or her ability to engage in healthy eating and physical activity behaviors. Specific modifications of the original measure included wording changes that revised items from being “exercise-specific” to include items relevant to healthy eating and physical activity. The current investigation looks to further validate the SE-HEPA using an exploratory factor analysis (EFA) in a confirmatory factor analysis (CFA) framework (Asparouhov & Muthén, 2009). This method was deemed appropriate as the simple structure assumption of CFA (i.e., all cross-loadings are zero) can be seen as too strict an assumption for this stage of measurement development. The EFA in CFA framework also allows for an investigation of items that may have significant cross-loadings. Thus, the current investigation was designed to evaluate the following hypotheses:

It was hypothesized that the SE-HEPA would evidence adequate reliability. Specifically, Cronbach’s alphas for the measure and each subscale (i.e., healthy eating, physical activity) would meet the minimum reliability criterion for internal consistency (>.70; Nunnally & Bernstein, 1994; Pedhazur & Schmelkin, 1991.

It was hypothesized that the SE-HEPA would evidence adequate construct validity. Specifically, because the measure was designed to include two separate constructs (i.e., healthy eating, physical activity), it was hypothesized that a two-factor model would be the best fit for the measure. Similarly, because the measure was designed to capture both healthy eating– and physical activity– specific self-efficacy, it was hypothesized that cross-loadings between the healthy eating factor and the physical activity factor will be nonsignificant; thus indicating that the SE-HEPA can be viewed as separate factors.

Because prior research indicates that dietary- and physical activity–specific self-efficacies are significantly related to weight status (e.g., Steele, Daratha, Bindler, & Power, in press), it was hypothesized that the SE-HEPA would evidence criterion-related validity, such that higher body mass index (BMI) percentile scores would be correlated with lower scores on the SE-HEPA factors.

Method

Participants and Procedures

With the approval of the Human Subjects Institutional Review Board at the appropriate institutions, all sixth, seventh, and eighth graders at a large middle school located in a suburban city in the Inland Northwest were given an informed consent letter to take home to their parents. With the approval of the parents, preadolescents were then invited to participate in the study. Students, whose parents consented, completed the questionnaires during their regular class time. Those students whose parents did not give permission, or students that declined to participate, were given an alternative assignment during the class. A total of 330 participated in the research, with 6 of these students turning in the measure blank and another 5 students failing to complete the entire measure. For the 319 students with complete surveys, 108 were sixth graders, 115 were seventh graders, and 96 were eighth graders.The sample of 319 participants consisted of slightly more male students (52%) and was approximately 80% White, 10% American Indian/Alaskan Native, 5% Hispanic, 4% Black, and 1% other. At the time of the study, 27% of the children registered at the school were eligible for free or reduced price lunch.

Measures

Demographic and anthropometric information

Students were asked to provide information pertaining to their age, gender, ethnicity, height, and weight. Because research has shown that adolescent self-reported height and weight is an accurate method for evaluating weight status (i.e., when compared with anthropomorphic measurement; Himes, Hannan, Wall, & Neumark-Sztainer, 2005), anthropometric data were gathered via self-report. Child BMI percentile scores were then calculated after referring to the Centers for Disease Control and Prevention growth charts containing age-specific median, standard deviation, and distribution skewness correction information (Kuczmarski et al., 2000). Researchers then calculated a raw BMI score for each child and categorized students based on weight status for subsequent analyses. That is, students with a BMI percentile for age and sex ≤85th percentile, but <95th percentile were classified as overweight and those with a BMI ≥95th percentile for age and sex were classified as obese.

Self-efficacy for healthy eating and physical activity (SE-HEPA)

As previously mentioned, the SE-HEPA measure was developed by Steele et al. (2008) and was based on Motl et al.s’ (2000) self-efficacy measure for exercise with adolescent females. The SE-HEPA contains eight items to measure self-efficacy for physical activity and eight items to measure self-efficacy for healthy eating. Each item was rated on a 5-point scale (i.e., disagree a lot, somewhat disagree, neither agree nor disagree, somewhat agree, agree a lot). Table 1 shows items for the SE-HEPA. In terms of the current investigation, it is important to note that each student was provided with a definition of physical activity and healthy eating immediately prior to the administration of the SE-HEPA to improve the likelihood that respondents would each have a basic understanding of the concepts related to the questionnaire. These definitions were based on recommendations from the Food and Drug Administration (U.S. Department of Health and Human Services, 2000) and the Rates of Perceived Exertion Scale (Borg, 1998). An initial validation study was conducted among a sample of 125 preadolescents from a middle school in a suburban city in the Inland Northwest that included students described as American Indian/Alaskan Native American (1%), Asian or Pacific Islander (9%), Black/non-Hispanic (3%), Hispanic (5%), multicultural (1%), and White/non-Hispanic (81%). An EFA indicated good fit for a two-factor model, S-B χ²(89) = 142.110, p = .003, comparative fit index (CFI) = .884, standardized root mean square residual (SRMR) = .059, and root mean square error of approximation (RMSEA) = .070. The Satorra–Bentler scaled chi-square (S-B χ²) statistics comparison revealed that the two-factor model provided significantly better fit than a one-factor model, S-Bχ²(15) = 82.555, p < .01, with factor loadings ranging from 0.123 to 0.763.

Table 1.

Self-Efficacy for Healthy Eating and Physical Activity (SE-HEPA) Items.

Physical activity items

1. I can be physically active during my free time on most days

2. I can ask my parent or other adults do physical activity things with me

3. I can be physically active during my free time on most days even if I could watch TV or play video games instead

4. I can be physically active during my free time on most days even if it is very hot or cold outside

5. I can ask my best friend to be physically active with me during my free time on most days

6. I can be physically active during my free time on most days even if I have to stay at home

7. I have the coordination I need to be physically active during my free time on most days

8. I can be physically active during my free time on most days no matter how busy my day is

Healthy eating items

1. I can eat healthy on most days

2. I can ask my parent or other adult for healthy foods

3. I can eat healthy foods even when unhealthy foods are available

4. I can eat healthy on most days even if I do not like what is available at home or school

5. I can ask my best friend to eat healthy with me on most days

6. I can eat healthy on most days even if I have to stay home

7. I know how to find and prepare healthy foods

8. I can eat healthy on most days no matter how busy my days is

Results

Item and Reliability Analyses

Table 2 shows the means, standard deviations, skewness, and kurtosis values for the 16 items. The preadolescents in this sample reported high levels of SE-HEPA, with most of the item means being slightly less than 4 (somewhat agree) on the 5-point scale. The mean BMI score for the preadolescents was 19.410 (SD = 3.179). In terms of weight status, approximately 5% of the participants were Underweight, 78% were “healthy weight,” 14% were “overweight,” and 3% were could be classified as “obese.” In terms of the reliability of the measure, consistent with hypotheses, internal consistency analyses indicated adequate reliability (i.e., >.70; Nunnally & Bernstein, 1994; Pedhazur & Schmelkin, 1991) for the full scale (α = .88), the healthy eating subscale (α = .85), and the physical activity subscale (α = .81) of the SE-HEPA.

Table 2.

Descriptive Information for Items.

Item	Mean	SD	Skewness	Kurtosis
Physical Activity 1	4.08	0.849	−0.523	−0.569
Physical Activity 2	3.49	1.223	−0.386	−0.969
Physical Activity 3	3.81	0.933	−0.459	−0.302
Physical Activity 4	3.90	0.946	−0.635	−0.142
Physical Activity 5	3.67	1.146	−0.576	−0.492
Physical Activity 6	3.91	1.026	−0.691	−0.272
Physical Activity 7	4.36	0.828	−1.386	2.039
Physical Activity 8	3.29	1.054	−0.056	−0.651
Healthy Eating 1	4.29	0.843	−1.320	2.193
Healthy Eating 2	4.43	0.851	−1.513	1.971
Healthy Eating 3	3.90	0.978	−0.738	0.131
Healthy Eating 4	3.80	0.974	−0.670	0.202
Healthy Eating 5	3.17	1.189	−0.190	−0.751
Healthy Eating 6	4.26	0.921	−1.290	1.337
Healthy Eating 7	4.39	0.877	−1.554	2.135
Healthy Eating 8	4.22	0.975	−1.180	0.710

Factor Analysis

Mplus 6.0 (Muthén & Muthén, 2010) was used to perform the EFA in the CFA framework (robust maximum likelihood estimation, maximum likelihood relaxation [MLR] with Geomin oblique rotation). MLR estimation was used with Geomin oblique rotation. The two-factor model provided a good fit in a global sense (i.e., CFI = .938, RMSEA = .054, and SRMR = .043) with the two-factor model providing the best fit to the data (p < .0001). Table 3 shows the completely standardized loadings for the two-factor model. Physical Activity 1 had a loading of 0.75 on its factor and a loading of −0.052 on the other factor, indicating good discriminant validity.

Table 3.

Factor Loadings for Total Sample in Exploratory Factor Analysis in Confirmatory Factor Analysis Framework: Total Items.

Item	Physical Activity Factor	Healthy Eating Factor
Physical Activity 1	0.750 (0.071)**	−0.052 (0.088)
Physical Activity 2	0.267 (0.097)**	0.323 (0.091)**
Physical Activity 3	0.667 (0.053)**	0.070 (0.070)
Physical Activity 4	0.675 (0.046)**	−0.005 (0.036)
Physical Activity 5	0.444 (0.079)**	0.115 (0.076)
Physical Activity 6	0.552 (0.080)**	0.021 (0.816)
Physical Activity 7	0.461 (0.093)**	0.119 (0.080)
Physical Activity 8	0.625 (0.053)**	−0.004 (0.037)
Healthy Eating 1	−0.115 (0.097)	0.786 (0.066)**
Healthy Eating 2	−0.002 (0.011)	0.660 (0.047)**
Healthy Eating 3	0.088 (0.088)	0.629 (0.071)**
Healthy Eating 4	0.112 (0.076)	0.580 (0.060)**
Healthy Eating 5	0.232 (0.011)*	0.322 (0.087)**
Healthy Eating 6	−0.031 (0.086)	0.792 (0.058)**
Healthy Eating 7	0.117 (0.080)	0.540 (0.081)**
Healthy Eating 8	0.036 (0.082)	0.658 (0.070)**

p < .05. **p < .01. ***p < .001.

Two of the items did not load on either factor. Specifically, Item 5’s loading was 0.322 on the healthy eating factor and 0.232 on the physical activity factor. Similarly, Item 2 on the physical activity scale also failed to load on either the healthy eating or physical activity factors, with loadings 0.267 and 0.323, respectively. The other items showed reasonable to good convergent and discriminant validity (see Table 3). The content of these two problematic items was inspected in an attempt to understand the reasons for the unexpected loadings. Item 2 for both scales asks about the respondent’s belief in his or her ability to engage an adult in physical activity/healthy eating practices with him or her. Item 5 for both scales are nearly identical asking about the respondent’s belief in his or her ability to engage a friend in physical activity/healthy eating practices with him or her. These items are unique in that they ask the respondent about engaging another person in healthy behaviors, rather than the respondent’s perceived ability to perform tasks individually (e.g., “I know how to find and prepare healthy foods”; “I can be physically active during my free time on most days.”). Based on this determination, Item 2 and Item 5 were dropped from both scales, resulting in parallel measures containing six items instead of the original eight. The analyses were repeated with the reduced scale (i.e., six healthy eating items, six physical activity items). For the reduced scale (i.e., six items per subscale), the two-factor model provide an excellent global fit (i.e., CFI = .976, RMSEA = .040, and SRMR = .030), with the fit of the two-factor being significantly better than the one-factor model (p < .0001).

Table 4 shows the completely standardized item-factor loadings for the reduced item set. Each of the physical activity items had a substantial (i.e., 0.476 to 0.751) loading on the physical activity factor with all the cross-loadings on the healthy eating factor being nonsignificant. Similar results occurred for the healthy eating items. Here, each healthy eating item had a substantial (i.e., 0.574 to 0.794) loading on the healthy eating factor with all the cross-loadings on the physical activity factor being nonsignificant. The correlation between the two factors was 0.508 (p < .001), indicating that the self-efficacy for healthy eating and physical activity could be viewed as separate factors. Based on the aforementioned criteria, a correlation of .51 would be considered an indication of good discriminant validity between the factors, since the factors only shared 25% of their variance.

Table 4.

Factor Loadings for Total Sample in Exploratory Factor Analysis in Confirmatory Factor Analysis Framework: Reduced Items.

Item	Physical Activity Factor	Healthy Eating Factor
Physical Activity 1	0.751 (0.074)**	−0.057 (0.089)
Physical Activity 3	0.697 (0.055)**	0.060 (0.073)
Physical Activity 4	0.678 (0.042)**	0.001 (0.022)
Physical Activity 6	0.555 (0.082)**	0.018 (0.093)
Physical Activity 7	0.476 (0.088)**	0.122 (0.093)
Physical Activity 8	0.623 (0.057)**	−0.008 (0.051)
Healthy Eating 1	−0.094 (0.083)	0.794** (0.058)
Healthy Eating 3	0.114 (0.111)	0.590** (0.090)
Healthy Eating 4	0.116 (0.102)	0.574** (0.077)
Healthy Eating 6	−0.020 (0.046)	0.784** (0.038)
Healthy Eating 7	0.115 (0.107)	0.535** (0.099)
Healthy Eating 8	0.042 (0.087)	0.675** (0.075)

p < .05. **p < .01. ***p < .001.

Relation of BMI to the Physical Activity and Healthy Eating Factors

To assess the criterion-related validity of the measure, bivariate correlations were conducted for the two self-efficacy factors from the SE-HEPA and BMI scores. Specifically, the total scores for the physical activity– and healthy eating–specific scales or the SE-HEPA were entered into a correlation matrix with total BMI percentile scores. The correlation of the BMI scores with the physical activity and healthy eating factors was −0.198 (p = .011) and −0.205 (p = .022), respectively. That is, preadolescents with higher BMI values thus reported lower self-efficacy for physical activity and healthy eating behaviors.

Discussion

This study was designed to further validate the SE-HEPA, a measure of self-efficacy for physical activity and healthy eating in children and adolescents. The SE-HEPA was adapted from Motl et al.’s (2000) exercise-specific measure of self-efficacy designed for adolescent females. Because healthy eating is also a critical component to pediatric weight loss, this measure had been adapted to include a parallel measure of healthy eating (Steele et al., 2008). The overarching goal of the current study was to test the SE-HEPA’s construct validity as a measure of self-efficacy for physical activity and healthy eating as well as the criterion-related validity by examining associations between the two self-efficacy factors and student BMI.

The first hypothesis, which sought to evaluate the reliability of the instrument, indicated that the SE-HEPA and each subscale evidenced adequate reliability based on standard reliability criteria (i.e., >.70 Nunnally & Bernstein, 1994; Pedhazur & Schmelkin, 1991). In terms of the second hypothesis, the fit of the two-factor model was supported. These analyses revealed that the two-factor model provided good global fit for the total sample. On reviewing the item-factor loadings, the results indicated certain items exhibited poor discriminant validity (i.e., Item 5 on the healthy eating scale, Item 2 on the physical activity scale). Because these two items were the only questions that asked the respondent about engaging another person in healthy behaviors, it may be that these items were assessing other influences on behavior, such as relational factors. That is, these two items may be tapping into a construct that is uniquely different from the construct that is being captured by the other items on the scale. For example, previous research has shown that primary sources of support (e.g., family and friends) are related to healthy eating practices among adolescents (Stanton, Green, & Fries, 2007). Similar studies have also found that various factors, such as peer victimization, parental distress, and gender differences all play a role in weight loss behaviors (Gray, Janicke, Ingerski, & Silverstein, 2008; Trinh, Rhodes, & Ryan, 2008). Because these items performed poorly in the study, these items were ultimately dropped from the questionnaire.

The second hypothesis, stating that the two-factor model (physical activity factor, healthy eating factor) would provide better fit than the one-factor model (self-efficacy), was also supported. The two-factor model provided a significantly better fit (ps < .001) with good global fit indices and no significant problems with the primary item-factor loadings. However, future research related to the measure should focus on replicating these findings among different populations. Specifically, an important next step will be to examine the factor structure of the measure among male-only, female-only, and ethnically/racially diverse samples to determine if the model used in the current investigation is further supported.

The third hypothesis, which evaluated the criterion-related validity for the scales of the SE-HEPA, was also supported. An examination of the primary correlations between the factors and BMI indicated that higher scores on self-efficacy were associated with lower self-reported BMI percentile scores. This result not only replicates previous findings (Steele et al., 2009) using only self-report measures of BMI but also replicates previous findings where anthropomorphic data were collected using actual height and weight measurements (Steele et al., 2008; Steele et al., in press).

Limitations, Future Research, and Clinical Applications

Certain limitations should be considered when interpreting the results of the current study. First, the investigation adopted a cross-sectional design. Thus, an assessment of the test–retest reliability would be important to better understand the stability of the measure over time. Of course, the dynamic nature of this particular construct would make it necessary to examine test–retest results within the context of other important variables (e.g., changes in weight status, stages of change). Second, no measures of actual physical activity and/or dietary intake were included. Future research that examines the relationships between physical activity– and healthy eating–specific self-efficacy and actual physical activity/dietary intake will be vital to this line of research, and would improve our understanding of the psychometric rigor of the SE-HEPA. The lack of ethnic, racial, and weight status (e.g., extreme obesity) diversity among the sample can be seen as a potential threat to the external validity of the study. Additionally, a larger sample size would allow for invariance testing among certain groups (e.g., males and females) to further validate the instrument. Finally, the measure of child weight status was captured via self-report from each student. Despite the fact that research has shown self-report to be a valid method for measuring height and weight among adolescents (Himes, Hannan, Wall, & Neumark-Sztainer, 2005), more sophisticated measures (e.g., dual-energy X-ray absorptiometry scan) would help improve our understanding of the relationship between self-efficacy and body composition.

Ultimately, having a valid measure of self-efficacy for behaviors that promote weight loss, such as the SE-HEPA, can have important clinical implications. Tailoring treatment efforts (e.g., a stepped-care approach; Carels et al., 2007) to match participating child and adolescent levels of self-efficacy could help improve adherence, and subsequently, treatment outcomes. For example, if a pretreatment assessment indicates a child/adolescent is exhibiting low levels of self-efficacy for healthy eating, but not physical activity, treatment strategies may be employed to help increase the child/ adolescent’s confidence in their ability to eat healthy. Specifically, motivational interviewing techniques, which have been shown to enhance self-efficacy for various health behavior (Carels et al., 2007), could be used to improve self-efficacy levels, which subsequently may influence adherence to treatment recommendations related to dietary changes. Again, improving adherence among children and adolescents participating in pediatric obesity intervention (e.g., family-based behavioral intervention) could ultimately improve health outcomes for this particular population.

Footnotes

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

References

Alinia

Hels

Tetens

(2009). The potential association between fruit intake and body weight: A review. Obesity Reviews, 10, 639-647.

Asparouhov

Muthén

(2009). Exploratory structural equation modeling. Structural Equation Modeling, 16, 397-438.

Bandura

(1977). Self-efficacy: Toward a unifying theory of behavior change. Psychological Review, 84, 191-215.

Bandura

(2005). Guide for constructing self-efficacy scales. In Pajares

Urdan

T. C.

(Eds.), Self-efficacy beliefs of adolescents (pp. 307-339). Charlotte, NC: Information Age.

Biddle

S. J. H.

Gorely

Stensel

D. J.

(2004). Health-enhancing physical activity and sedentary behaviour in children and adolescents. Journal of Sports Sciences, 22, 679-701.

Borg

(1998). Borg’s perceived exertion and pain scales. Champaign, IL: Human Kinetics.

Burke

Arkowitz

Menchola

(2003). The efficacy of motivational interviewing: A meta-analysis of controlled clinical trials. Journal of Consulting and Clinical Psychology, 71, 843-861.

Carels

R. A.

Darby

Cacciapaglia

H. M.

Konrad

Coit

Harper

Versland

(2007). Using motivational interviewing as a supplement to obesity treatment: A stepped-care approach. Health Psychology, 26, 369-374.

Dallow

C. B.

Anderson

(2003). Using self-efficacy and a transtheoretical model to develop a physical activity intervention for obese women. American Journal of Health Promotion, 17, 373-381.

10.

Davidson

K. K.

Marshall

S. J.

Birch

L. L.

(2006). Cross-sectional and longitudinal associations between T.V. viewing and girl’s body mass index, overweight status, and percentage of body fat. Journal of Pediatrics, 149, 32-37.

11.

Gray

W. N.

Janicke

D. M.

Ingerski

L. M.

Silverstein

J. H.

(2008). The impact of peer vicitization, parent distress and child depression on barrier formation and physical activity in overweight youth. Journal of Developmental & Behavioral Pediatrics, 29, 26-33.

12.

Himes

Hannan

Wall

Neumark-Sztainer

(2005). Factors associated with errors in self-reports of stature, weight, and Body Mass Index in Minnesota adolescents. Annals of Epidemiology, 15, 272-278.

13.

Jelalian

Markov-Wember

Bungeroth

Burmaher

(2007). Practitioner review: Bridging the gap between research and clinical practice in pediatric obesity. Journal of Child Psychology and Psychiatry, 48, 115-127.

14.

Kelder

S. H.

Perry

C. L.

Klepp

K. I.

Lytle

L. L.

(1994). Longitudinal tracking of adolescent smoking, physical activity, and food choice behaviors. American Journal of Public Health, 84, 1121-1126.

15.

Kelly

(1967). Attribution theory in social psychology. Nebraska Symposium on Motivation, 15, 192-238.

16.

Kuczmarski

R. J.

Ogden

C. L.

Grummer-Strawn

L. M.

Flegal

K. M.

Guo

S. S.

Wei

. . . Johnson

C. L.

(2000). CDC growth charts: United States. Advance Data, 314, 1-27. Retrieved from http://www.cdc.gov/nchs/data/ad/ad314.pdf

17.

Linde

J. A.

Rothman

A. J.

Baldwin

A. S.

Jeffery

R. W.

(2006). The impact of self-efficacy on behavior change and weight change among overweight participants in a weight loss trial. Health Psychology, 25, 282-291.

18.

Motl

R. W.

Dishman

R. K.

Trost

S. G.

Saunders

R. P.

Dowda

Felton

. . . Pate

P. R.

(2000). Factorial validity and invariance of questionnaires measuring social cognitive determinants of physical activity among adolescent girls. Preventive Medicine, 31, 584-594.

19.

Muthén

L. K.

Muthén

B. O.

(2010). Mplus statistical analysis with latent variables user’s guide (Version 6.0). Los Angeles, CA: Muthén & Muthén.

20.

Nunnally

J. C.

Bernstein

I. R.

(1994). Psychometric theory. New York, NY: McGraw-Hill.

21.

Ogden

C. L.

Carroll

M. D.

Flegal

K. M.

(2003). Epidemiologic trends in overweight and obesity. Endocrinology and Metabolism Clinics of North America, 32, 741-760.

22.

Ogden

C. L.

Carroll

M. D.

Flegal

K. M.

(2008). High body mass index for age among US children and adolescents, 2003-2006. Journal of the American Medical Association, 299, 2401-2405.

23.

Ogden

C. L.

Flegal

K. M.

Carroll

M. D.

Johnson

C. L.

(2002). Prevalence and trends in overweight among US children and adolescents, 1999-2000. Journal of the American Medical Association, 288, 1728-1732.

24.

Ogden

C. L.

Yanovski

S. Z.

Carroll

M. D.

Flegal

K. M.

(2007). The epidemiology of obesity. Gastroenterology, 132, 2087-2102.

25.

Pedhazur

E. J.

Schmelkin

L. P.

(1991). Measurement, design, and analysis: An integrated approach. Mahwah, NJ: Erlbaum.

26.

Prochaska

J. O.

DiClemente

C. C.

(1983). Stages and processes of self-change of smoking: Toward an integrative model of change. Journal of Consulting and Clinical Psychology, 51, 390-395.

27.

Raynor

Maier

(2006). The childhood obesity epidemic: Key eating and activity behaviors to address in treatment. Medicine and Health, Rhode Island, 89, 241-243, 246.

28.

Spear

B. A.

Barlow

S. E.

Ervin

Ludwig

D. S.

Saelens

B. E.

Schetzina

K. E.

Taveras

E. M.

(2007). Recommendations for treatment of child and adolescent overweight and obesity. Pediatrics, 120(Suppl. 4), S254-S288.

29.

Stanton

C. A.

Green

S. L.

Fries

E. A.

(2007). Diet-specific social support among rural adolescents. Journal of Nutrition Education and Behavior, 39, 214-218.

30.

Steele

M. M.

Bindler

Power

T. G.

Daratha

(2008, April). Measures of self-determination and self-efficacy in pre-adolescents. Poster session presented at the annual Western Institute of Nursing conference, Garden Grove, CA.

31.

Steele

M. M.

Daratha

K. B.

Bindler

Power

T. M.

(in press). Predictors of weight status: Examining the relationship between self-efficacy and clinical indicators of health in pre-adolescents. Health Education & Behavior.

32.

Steele

M. M.

Steele

R. G.

Hunter

(2009). Family adherence as a predictor of child outcome in an intervention for pediatric obesity: Different outcomes for self-report and objective measures. Children’s Health Care, 38, 64-75.

33.

Trinh

Rhodes

R. E.

Ryan

S. M.

(2008). Gender differences in belief-based targets for physical activity interventions among adolescents. Social Behavior and Personality, 36, 77-86.

34.

U.S. Department of Health and Human Services. (2000). Healthy people 2010: Understanding and improving health. Retrieved from http://www.healthypeople.gov/2010/Document/tableofcontents.htm#under

35.

Wiltink

Dippel

Szczepanski

Thiede

Alt

Beutel

M. E.

(2007). Long-term weight loss maintenance after inpatient psychotherapy of severely obese patients based on randomized study: Predictors and maintaining factors of health behavior. Journal of Psychosomatic Research, 62, 691-698.