This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in the Journal of Medical Internet Research, is properly cited. The complete bibliographic information, a link to the original publication on http://www.jmir.org/, as well as this copyright and license information must be included.
Technology-mediated obesity treatments are commonly affected by poor long-term adherence. Supportive Accountability Theory suggests that the provision of social support and oversight toward goals may help to maintain adherence in technology-mediated treatments. However, no tool exists to measure the construct of supportive accountability.
This study aimed to develop and psychometrically validate a supportive accountability measure (SAM) by examining its performance in technology-mediated obesity treatment.
Secondary data analyses were conducted in 2 obesity treatment studies to validate the SAM (20 items). Study 1 examined reliability, criterion validity, and construct validity using an exploratory factor analysis in individuals seeking obesity treatment. Study 2 examined the construct validity of SAM in technology-mediated interventions involving different self-monitoring tools and varying amounts of phone-based interventionist support. Participants received traditional self-monitoring tools (standard, in-home self-monitoring scale [SC group]), newer, technology-based self-monitoring tools (TECH group), or these newer technology tools plus additional phone-based support (TECH+PHONE group). Given that the TECH+PHONE group involves more interventionist support, we hypothesized that this group would have greater supportive accountability than the other 2 arms.
In Study 1 (n=353), the SAM showed strong reliability (Cronbach α=.92). A factor analysis revealed a 3-factor solution (representing Support for Healthy Eating Habits, Support for Exercise Habits, and Perceptions of Accountability) that explained 69% of the variance. Convergent validity was established using items from the motivation for weight loss scale, specifically the social regulation subscale (r=0.33;
The SAM has strong reliability and validity across the 2 studies. Future studies may consider using the SAM in technology-mediated weight loss treatment to better understand whether support and accountability are adequately represented and how supportive accountability impacts treatment adherence and outcomes.
ClinicalTrials.gov NCT01999244; https://clinicaltrials.gov/ct2/show/NCT01999244
Technology-mediated health behavior change interventions have become ubiquitous [
Given this challenge, research has focused on methods to improve adherence to technology-based behavioral interventions. To date, one of the most effective strategies has been the provision of additional support (eg, via telephone, email, text messages or via smartphone apps) from interventionists or lifestyle coaches [
Consistent with these empirical findings, Mohr et al [
The objective of this study was to describe the development and validation of a measure for assessing supportive accountability within the context of technology-mediated programs for the treatment of adult obesity. Building off of the Social Support for Diet and Exercise Behaviors Scale by Sallis et al [
The SAM was developed by adding novel items representing the construct of accountability to items from the existing Social Support for Diet and Exercise Behaviors Scale developed by Sallis et al [
To ensure balance in the number of items, 10 items were selected from the Sallis Social Support measure [
A total score for the SAM was created using the 10 support items and 10 accountability items. The Social Support subscale involves summing all social support items. Since the accountability items were measured on a 7-point scale and the social support items on a 5-point scale, the scores for the accountability measure involved summing all accountability items, multiplying the sum by 5, and then dividing by 7 to yield identical maximum scores for the 2 subscales. The total scores for social support and accountability were then summed to create an overall SAM score, with higher scores indicating higher levels of supportive accountability (possible range of total SAM scores 17.14-100). We then examined the psychometric characteristics of the SAM across 2 weight management trials (the study details are provided below under the Methods section for Study 1 and Study 2). Study procedures for Study 1 and Study 2 were approved by The Miriam Hospital Institutional Review Board, and secondary analyses of Study 2 data were approved by the University of Florida Institutional Review Board.
Study 1 evaluated the psychometric characteristics of the SAM in adults with obesity enrolled in a behavioral weight management trial. Only baseline data from this trial were used, allowing for the evaluation of the SAM in the relevant population before any intervention or support.
Study 1 included 353 adults (aged 40-60 years) with obesity (BMI between 30 kg/m2 and 40 kg/m2). Full eligibility criteria and recruitment procedures have been described elsewhere [
All measures were collected at baseline.
Standard demographics (eg, sex, age, race/ethnicity) were collected via a survey.
Height and weight were objectively assessed, with participants wearing one layer of light indoor clothing and shoes removed.
The SAM was used to assess supportive accountability (see the Supportive Accountability Measure Development section for details).
The Treatment Self-Regulation Questionnaire (TSRQ) was administered to assess its convergent validity with the SAM. Given that the SAM is expected to measure social accountability, we examined whether the social external regulation subscale of the TSRQ [
Social pressure was assessed using an item from the Motivating Factors for Weight Loss Scale, developed to assess motivation for weight loss among participants in the National Weight Control Registry [
Analyses were conducted using SPSS Statistics for Windows, version 25 (IBM Corp). The reliability of SAM items was assessed via the Cronbach alpha [
Study 2 was a randomized weight management trial investigating the impact of newer self-monitoring technology (ie, a Bluetooth-enabled activity monitor, a smart scale, and a website/smartphone app that synced with both of these devices and allowed individuals to self-monitor caloric intake) and phone coaching on weight loss. It was hypothesized that participants who received interventionist support through phone coaching would report significantly higher supportive accountability, as assessed by the SAM, at a 6-month posttest because of the presence of additional support. Moreover, we hypothesized that higher supportive accountability at intervention posttest would be associated with greater intervention adherence. As an exploratory aim, we investigated whether higher supportive accountability at the posttest was associated with greater weight loss from baseline to posttest.
Study 2 participants were 80 adults (aged between 18 and 70 years) with overweight or obesity (BMI between 27 kg/m2 and 40 kg/m2) who reported having access to a computer and Wi-Fi at home [
Study 2 was a randomized trial that examined the impact of a 6-month weight loss intervention in which participants were randomized to 1 of 3 treatment groups, using traditional self-monitoring tools (a paper food record, a printed calorie reference book, a standard pedometer, and a standard in-home scale—SC group), newer, technology-based self-monitoring tools (Fitbit Zip, Fitbit Aria smart scale, and use of the Fitbit app/website to track dietary intake—TECH group), or these newer technology tools plus phone-based interventionist support (TECH+PHONE group). All participants received a one-time, group-based Weight Loss 101 session that provided information on how to accurately monitor calories, weight, and physical activity and weight management goals for calories, exercise, and weight loss. Participants were also taught how to use their assigned self-monitoring tools. Participants randomized to SC (n=26) and TECH (n=27) received self-monitoring tools only; they did not receive any interventionist support. Participants randomized to TECH+PHONE (n=27) received the additional phone-based intervention (8 weekly, 4 biweekly, and 2 monthly contacts; each lasted 10-15 min), delivered by trained interventionists (either a clinical psychologist or dietitian, both experienced in delivering behavioral weight management programs), using a manualized protocol.
Standard demographics (eg, sex, age, race/ethnicity) were collected via a survey at baseline.
Height and weight were measured with participants wearing light indoor clothing and with shoes removed. Height was measured at baseline and weight was measured at baseline and at the 6-month posttest. Weight change was operationalized as percent weight loss from baseline to the posttest visit.
Supportive accountability was assessed at the 6-month posttest using the SAM.
Weight management strategies were assessed at the 6-month posttest using the Weight Control Strategies Scale (WCSS) [
Analyses were conducted using SPSS. Reliability was reassessed in this sample using Cronbach alpha. Construct validity of the SAM was assessed using a one-way analysis of variance, investigating differences in SAM scores by treatment group. We hypothesized that, at the 6-month posttest, the SAM score would be significantly higher in the TECH+PHONE condition than in the SC and TECH conditions, given that this condition was provided with additional interventionist support.
Study 1 included a total of 350 participants (
The overall reliability of the SAM was excellent, as demonstrated by the internal consistency (Cronbach α=.92). The lowest item to total SAM correlation (
Results from the EFA demonstrated that a three-factor solution provided the best fit (
Convergent and divergent validity analyses revealed statistically significant correlations between the SAM total score and TSRQ items representing external motivation for weight loss (
Baseline and demographic characteristics of participants in study 1
Characteristic | Study 1 (n=350) | |
Age (years), mean (SD) | 51.7 (5.6) | |
BMI (kg/m2), mean (SD) | 34.8 (3.3) | |
|
||
|
Male | 80 (22.9) |
|
Female | 270 (77.1) |
|
||
|
High school or less | 38 (10.9) |
|
Vocational training | 24 (6.9) |
|
Some college | 83 (23.7) |
|
College degree | 108 (30.9) |
|
Graduate degree | 97 (27.7) |
|
||
|
American Indian or Alaska Native | 6 (1.7) |
|
Asian | 1 (0.3) |
|
Black or African American | 36 (10.3) |
|
White | 262 (74.9) |
|
Other | 36 (10.3) |
|
||
|
Hispanic or Latino | 45 (12.9) |
|
Not Hispanic or Latino | 304 (86.9) |
Scree plot for exploratory factor analysis of the supportive accountability measure.
Three-factor structure for Supportive Accountability Measure items.
Items | Support for Healthy Eating | Support for Exercise Habits | Perceptions of Accountability | |
|
||||
|
1. Encouraged me not to eat |
0.81 | N/Aa | N/A |
|
2. Discussed my eating habit changes with me (asked me how I’m doing with my eating changes) | 0.83 | N/A | N/A |
|
3. Reminded me not to eat high fat, high calorie foods | 0.85 | N/A | N/A |
|
4. Complimented me on changing my eating habits (“Keep it up. We are proud of you.”) | 0.68 | N/A | N/A |
|
5. Commented if I went back to my old eating habits | 0.71 | N/A | N/A |
|
6. Exercised with me | N/A | 0.88 | N/A |
|
7. Offered to exercise with me | N/A | 0.87 | N/A |
|
8. Gave me helpful reminders to exercise (“Are you going to exercise tonight?”) | N/A | 0.85 | N/A |
|
9. Gave me encouragement to stick with my exercise program | N/A | 0.78 | N/A |
|
10. Changed their schedule so we could exercise together | N/A | 0.81 | N/A |
|
11. I feel accountable to others (eg, friends, family, or doctor) for meeting my weight goals. | N/A | N/A | 0.77 |
|
12. I feel accountable to others (eg, friends, family, or doctor) for meeting my dietary goals. | N/A | N/A | 0.79 |
|
13. I feel accountable to others (eg, friends, family, or doctor) for meeting my exercise goals. | N/A | N/A | 0.79 |
|
14. I feel that I would let others down (eg, friends, family, or doctor) if I did not meet my weight goals. | N/A | N/A | 0.93 |
|
15. I feel that I would let others down (eg, friends, family, or doctor) if I did not meet my dietary goals. | N/A | N/A | 0.94 |
|
16. I feel that I would let others down (eg, friends, family, or doctor) if I did not meet my exercise goals. | N/A | N/A | 0.92 |
|
17. Feeling accountable to others (eg, friends, family, or doctor) has helped me control my weight. | N/A | N/A | 0.57 |
|
18. Feeling accountable to others (eg, friends, family, or doctor) has helped me stay on track with my diet. | N/A | N/A | 0.55 |
|
19. Feeling accountable to others (eg, friends, family, or doctor) has helped me stay on track with my physical activity. | N/A | N/A | 0.54 |
|
20. In general, I feel accountable to others to control my weight. | N/A | N/A | 0.74 |
aN/A: not applicable.
Of the 80 participants included in Study 2, 55 completed the SAM at the 6-month assessment and were thus included in these analyses.
At the 6-month follow-up, the mean SAM score was 47.86 (SD 20.88); scores ranged from 17.14 to 94.57 (highest possible score of 100). Replicating the analyses conducted in Study 1, the reliability analyses for 6-month SAM scores demonstrated strong internal consistency (Cronbach α=.95).
Baseline and demographic characteristics of participants by group in study 2.
Characteristic | Study 2 (n=55) | |||
|
SCa (n=17) | TECHb (n=17) | TECH+PHONEc (n=21) | |
Age (years), mean (SD) | 54.9 (9.7) | 46.4 (12.7) | 52.9 (11.28) | |
BMI (kg/m2), mean (SD) | 34.0 (4.0) | 33.0 (3.4) | 32.2 (3.3) | |
|
||||
|
Male | 4 (23) | 2 (11) | 4 (19) |
|
Female | 13 (76) | 15 (88) | 17 (81) |
|
||||
|
High school or less | 1 (5) | 0 (0) | 2 (9) |
|
Vocational training | 0 (0) | 1 (5) | 0 (0) |
|
Some college | 4 (23) | 3 (17) | 5 (23) |
|
College degree | 6 (35) | 7 (41) | 6 (28) |
|
Graduate degree | 6 (35) | 6 (35) | 8 (38) |
|
||||
|
American Indian or Alaska Native | 0 (0) | 0 (0) | 0 (0) |
|
Asian | 0 (0) | 0 (0) | 0 (0) |
|
Black or African American | 2 (11) | 2 (11) | 0 (0) |
|
White | 15 (88) | 15 (88) | 21 (100) |
|
Other | 0 (0) | 0 (0) | 0 (0) |
|
||||
|
Hispanic or Latino | 0 (0) | 1 (6) | 1 (4) |
|
Not Hispanic or Latino | 17 (100) | 16 (94) | 20 (95) |
aSC: standard in-home scale group.
bTECH: technology-based self-monitoring tools group.
cTECH+PHONE: technology-based self-monitoring tools plus phone-based support group.
As hypothesized, there were significant differences in the 6-month total SAM scores by treatment group (
Between-group differences in SAM subscales were also examined (
Six-month scores on supportive accountability measure (SAM) and each of the three SAM subscales by intervention group.
Scale | Value, mean (SE) | ||
|
|||
|
SCa | 30.9 (3.8) | |
|
TECHb | 44.9 (3.8) | |
|
TECH+PHONEc | 64.0 (3.5) | |
|
|||
|
SC | 9.8 (1.2) | |
|
TECH | 14.5 (1.2) | |
|
TECH+PHONE | 20.0 (1.1) | |
|
|||
|
SC | 9.2 (1.4) | |
|
TECH | 14.0 (1.4) | |
|
TECH+PHONE | 13.7 (1.2) | |
|
|||
|
SC | 11.9 (2.6) | |
|
TECH | 16.3 (2.6) | |
|
TECH+PHONE | 30.3 (2.3) |
aSC: standard in-home scale group.
bTECH: technology-based self-monitoring tools group.
cTECH+PHONE: technology-based self-monitoring tools plus phone-based support group.
Correlation between the supportive accountability measure (total scores and subscales) and adherence to weight control strategies and weight change from baseline to 6-month posttest.
Scale | Total SAMa score | Support for Healthy Eating Habits | Support for Exercise | Perceptions of Accountability | |||||||||||
|
|
|
|
|
|||||||||||
|
0.47 | .003b | 0.26 | .06 | 0.23 | .09 | 0.50 | <.001b | |||||||
|
Dietary choices | 0.29 | .03b | 0.14 | .30 | 0.15 | .27 | 0.32 | .01b | ||||||
|
Self-monitoring strategies | 0.47 | .003b | 0.28 | .03b | 0.26 | .06 | 0.49 | .002b | ||||||
|
Physical activity | 0.23 | .08 | −0.02 | .90 | 0.15 | .28 | 0.31 | .02b | ||||||
|
Psychological coping | 0.45 | .007b | 0.36 | .006b | 0.16 | .25 | 0.46 | <.001b | ||||||
Percent weight change from baseline to 6 months | −0.26 | .06 | −0.21 | .11 | −0.12 | .39 | −0.25 | .06 |
aSAM: supportive accountability measure.
bStatistically significant (
Study 1 evaluated the psychometric properties of the novel, theory-based SAM in a sample of adults interested in a behavioral weight management trial. EFA revealed a three-factor solution for the SAM, representing subscales for Support for Healthy Eating Habits, Support for Exercise Habits, and Perceptions of Accountability. All items were retained, leaving 10 items representing social support and 10 items representing accountability, which together form the theoretical basis for the construct of supportive accountability [
Study 2 evaluated the criterion validity of the SAM. Consistent with the hypotheses, participants provided with additional phone support (TECH+PHONE) reported higher SAM scores at the end of a 6-month weight loss program compared with participants who did not receive phone support (TECH only or SC only). Interestingly, participants who did not receive phone support but were provided with newer, technology-based self-monitoring tools reported higher supportive accountability at the end of the intervention compared with participants randomized to self-monitoring using traditional tools (a standard pedometer, bathroom scale, a calorie reference book, and paper self-monitoring logs used to track physical activity, weight, and caloric intake). Subscale analyses revealed that this was likely driven by increased feelings of social support but not accountability. It may be that the brief, automated feedback provided by these tools gave participants a sense of support. Specifically, these tools provided immediate feedback to participants who met short- and long-term goals (such as notifications on the activity monitor and pushed smartphone notifications and
Consistent with the Supportive Accountability Theory [
Finally, consistent with Mohr’s theory that supportive accountability could increase intervention adherence [
Overall, data from 2 studies were used to examine the psychometric properties of the SAM, a new survey to assess supportive accountability for weight management behaviors. Across both studies, the SAM demonstrated excellent internal consistency and construct validity. This study has some limitations. Participants were predominantly female and non-Hispanic white, which limits the generalizability of the study results. Future research is needed to examine whether SAM demonstrates a similar factor structure, internal consistency, and validity in more diverse samples. The small sample size in Study 2, combined with the fact that participants in only 2 of the 3 groups were asked to use technology-based self-monitoring tools, precluded the investigation into whether the SAM was associated with objective engagement with the technology-based self-monitoring tools. This study also did not assess test-retest reliability or sensitivity to change over time. Future work in these areas would strengthen confidence in this measure for assessing the construct of supportive accountability and further provide important results that could inform future theory development. Finally, the scoring on this measure was complicated by the different scales used for scoring the accountability items (which used a 7-point scale) and the social support items (which used a 5-point scale). The accountability items were developed before the selection of social support items from the Sallis questionnaire, and a 7-point scale was chosen to optimize variability in responses. Future research should investigate whether this scale performs similarly when all items use the same scale (eg, either 5 or 7 response points).
This research also has notable strengths. The development of SAM was theory-based, relying on the Supportive Accountability Theory by Mohr [
Supportive accountability was developed as a construct within the context of technology-mediated
Moreover, although the construct of supportive accountability suggests the importance of human support [
Considering the ever-growing technological innovations, the SAM will help researchers better understand the factors that drive the effectiveness of technology-based treatments. This use of the SAM may thus guide the development of more effective interventions and help improve foundational knowledge regarding the mechanisms that drive treatment effects in technologically mediated treatment.
The supportive accountability measure and scoring.
Inter-item correlations between supportive accountability measure items.
CONSORT-eHEALTH checklist (V 1.6.2).
exploratory factor analysis
Kaiser-Meyer-Olkin
Supportive Accountability Measure
Treatment Self-Regulation Questionnaire
Weight Control Strategies Scale
None declared.