Health Education Research, Vol. 14, No. 2, 257-267,
April 1999
© 1999 Oxford University Press
A tailored multimedia nutrition education pilot program for low-income women receiving food assistance
Department of Nutrition,
1 Department of Health Behavior and Health Education, and
2 Carolina Population Center, University of North Carolina, Chapel Hill, NC 27599, and
3 People Designs, 1200 Broad Street, Durham, NC 27705, USA
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
This article describes the development and pilot evaluation of a tailored multimedia program to improve dietary behavior among 378 low-income women enrolled in the Food Stamp program in Durham, North Carolina. After randomization to intervention or control groups, participants completed a baseline survey and were resurveyed 1–3 months post-intervention. Measures included dietary fat intake assessed using a brief food-frequency questionnaire, stage of change, knowledge of low-fat foods, self-efficacy and eating behavior questions. The computer-based intervention consisted of a tailored soap opera and interactive `info-mercials' that provided individualized feedback about dietary fat intake, knowledge and strategies for lowering fat based on stage of change. At follow-up, intervention group participants had improved significantly in knowledge (P < 0.001), stage of change (P < 0.05) and certain eating behaviors (P < 0.05) compared to the control group. Both study groups had lowered their reported fat intake markedly at follow-up (P < 0.001), but did not differ significantly from each other. A majority of participants rated the program as very helpful and were interested in using a similar program in the future. The findings of this pilot study suggest that computerized tailored self-help health promotion programs may be effective educational interventions for lower income and minority populations.
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Effective nutrition education programs can enable individuals to improve their health and prevent diet-related chronic diseases. Dietary intervention programs for lower income and minority populations are particularly needed because these groups suffer a disproportionate burden of diet-related chronic disease morbidity and mortality (McGinnis and Foege, 1993
).
Health communications are more likely to be effective when they are culturally appropriate, and incorporate the relevant concerns, barriers and motivators of individuals (Witte, 1995). Recently, computer-based systems have been developed to tailor health information to participants' needs and interests (Skinner et al., 1993). In previous research, printed computer-tailored nutrition messages were shown to significantly reduce dietary fat intake among family practice patients and Dutch oil company employees (Campbell et al., 1994; Brug et al., 1996). The effectiveness of tailored dietary messages to lower dietary fat has not, however, been evaluated among predominantly low-income and minority populations.
This paper will describe the development and evaluation of StampSmart, a US Department of Agriculture-funded pilot project aimed at developing innovative nutrition education strategies for the Food Stamp program. The program targeted women because women and children represent the majority of Food Stamp recipients, and women are often the `gatekeepers' of food and nutrition for the family. Computer-tailored multimedia was used as the intervention strategy for several reasons. First, a stand-alone, self-help program provided a way to offer nutrition education at the Food Stamp office without burdening agency staff with extra responsibilities. Second, use of multimedia enabled the program to combine education with television-like entertainment in order to attract and hold the user's attention. Third, use of audio and visuals rather than printed messages made the intervention more accessible for participants with low literacy skills.
The conceptual framework for StampSmart (see Figure 1) was based on health behavior theories, including Social Cognitive Theory (Bandura, 1986) and the Stages-of-Change Transtheoretical Model (Prochaska and DiClemente, 1983; Prochaska et al., 1992). Social marketing principles were used to determine the perceived interests and concerns of the target audience in order to inform the program's content and format (Novelli, 1990). We conducted formative interviews with 54 female Food Stamp recipients to elicit cultural, personal and social factors associated with eating habits, barriers to participating in nutrition education programs and media preferences. Women said that they were concerned about their own and their children's nutrition but did not want to participate in nutrition education sessions because they would be `boring—like something out of the 70s'. Most women cited television as their major source of health information, and they preferred soap opera and drama plots over other types of programs. Based on this information, we developed a program that included a video soap opera, Sisters at Heart, interspersed with tailored `info-mercial' breaks. The program incorporated specific strategies to increase self-efficacy (confidence) regarding low-fat eating and advance individuals through the stages of change. These strategies included modeling of healthy behavior changes in the soap opera story, interactive exercises with immediate feedback regarding nutrition knowledge, stage-based behavioral messages and recommending small steps to initiate action.
|
Few studies have evaluated the impact of multimedia programs on targeted behavior changes (Maibach and Holtgrave, 1995
). This paper will report the pilot program's effects related to low-fat food choices, specifically (1) nutrition knowledge, (2) self-efficacy, (3) stage of change and (4) dietary fat intake. Process evaluation measures of participant satisfaction and perceived usefulness of the program are also reported. |
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Recruitment and data collection
Participants were recruited at the Food Stamp certification office in Durham, North Carolina from January through April, 1995. Trained research assistants assessed eligibility, obtained informed consent and assured women that participation did not affect Food Stamp eligibility. Women were eligible if they were 18 years of age or older, spoke English and either had children under 18 living at home or were pregnant.
All women who came to the certification office during the study period were invited to participate in the program. Of 2046 initial contacts, 413 people were excluded due to ineligibility. A total of 987 approaches met with refusal; however, many of these individuals participated on a subsequent visit. We were not able to determine the number of `true' refusers versus the number who initially refused but later agreed to participate, because we did not obtain names until participation was established. Lack of time was the leading reason given for refusing (660); other reasons included not being interested (194), not feeling well (11) or unspecified reason (122).
The baseline study sample consisted of 526 women selected from an initial pool of 644 individuals who consented and met eligibility criteria. A total of 118 people were excluded from the baseline sample for the following reasons: survey incomplete (100), completed survey twice (3) and survey data not matched to valid ID number (15). Women were randomized to control or intervention groups based on the day they participated, to decrease potential contamination. During pre-testing, we learned that women tended to come to the Food Stamp office with friends or relatives and therefore were likely to compare the information they received.
Participants were resurveyed 1–3 months post intervention to assess change. Telephone calling was initiated 1 month after the intervention; however, it took up to an additional 2 months to contact or exclude some participants due to the transience of this population. The final sample consisted of 378 women who completed both surveys (response rate of 72% of baseline participants). Nearly all were resurveyed by telephone (351 people); however, 27 people who could not be reached by telephone were resurveyed in person at the Food Stamp office. Interviewers were blinded to participants' study group membership. Reasons for non-response included: refusals (7), unable to leave a message or unable to complete the interview after at least eight attempts (87) and disconnected telephone or no telephone number provided (54). Respondents to the follow-up survey did not differ from non-respondents in demographic characteristics or dietary fat intake.
All baseline (T1) and immediate post-program or (for the control group) post-baseline survey (T2) data were collected in the Food Stamp certification office on the same day the woman agreed to participate. Intervention group participants completed the surveys using the interactive computer program and these data were automatically saved into a data file. To avoid biasing the intervention group's responses, the dietary and psychosocial questions were positioned at the beginning of the multimedia program before any nutrition education or feedback had occurred. For the control group, trained research assistants read each question aloud, and participants then marked their answers on survey forms using paper and pencil. The research assistants were trained to read the questions using the same pacing and voice inflections as the narrator in the multimedia program, in order to increase comparability of the two methods. Two different data collection methods were used because of resource constraints that prevented us from implementing a comparable `survey-only' computer program for controls. Telephone survey methods were used for follow-up because Food Stamp participants were not required to return for re-certification for 6–12 months and it was likely that many participants would have left the area or have been unreachable after that length of time. Follow-up telephone survey (T3) data were recorded onto paper surveys by trained interviewers. All paper survey data were double entered into data files and checked for accuracy against hard copies.
Measures
Demographic characteristics (T1)
Participants provided information at baseline about their age, race/ethnicity, education, the number and ages of their children, and whether they were currently pregnant or breastfeeding.
Stage of change (T1 and T3)
Participants were categorized into one of the following stages related to lowering dietary fat based on two questions derived from the Prochaska and DiClemente Transtheoretical framework (Prochaska et al., 1992). The first question asked participants to state whether or not they thought that most of the foods they ate were low in fat. Those who responded `yes' were categorized in the action/maintenance stage (these two stages were not separated). Those who answered `no' were then asked a question to determine their plans and interest in starting to eat foods low in fat. Those who were not thinking of starting were classified in the precontemplation stage, those who were thinking of starting were classified in the contemplation stage and those who were planning to start within the next month were classified in the preparation stage.
Self-efficacy (T1, T2 and T3)
One question about self-efficacy, `How sure are you that can eat foods low in fat over the next month?' was measured using a five-point Likert scale, with 5 = very sure and 1 = not at all sure. This question was adapted from an item assessing self-efficacy regarding fruit and vegetable consumption, which has been validated against multi-item self-efficacy scales, stage of change and dietary intake (Campbell et al., 1998) (R. Feldman, pers. commun.).
Knowledge (T1 and T3)
Six multiple choice items asked participants to choose the low-fat food from a set of options. Knowledge questions asked about low-fat breakfast foods, fast foods, snack foods and meal components (meat, starch and vegetable choices). Correct answers were then summed to create a low-fat knowledge score (possible range = 0–6).
Perceived overweight (T1)
Participants were asked whether or not they felt they needed to lose weight.
Autonomy (T1)
Responsibility for food shopping, planning meals and preparing meals was measured using three questions rated on a three-point scale, with 1 = little or none, 2 = about half and 3 = most or all (Beresford et al., 1992). High autonomy was defined as having most or all responsibility for all three functions.
Dietary fat score (T1 and T3)
A 16-item food-frequency questionnaire (FFQ) derived from two validated instruments was used to determine a dietary fat score (Block et al., 1986, 1989). Two additional items were included to assess consumption frequency of low-fat foods (skim/low-fat milk, pretzels or graham crackers) but were not factored into the overall score. Photographs of a medium serving of each item were provided to assist participants visually with the questions. Seven response choices were presented for each item: 3 or more/day, 2/day, about every day, 2–4/week, once a week, 1–3 times/month or never or almost never. Dietary fat scores were obtained by multiplying frequency of consumption adjusted to daily intake (3, 2, 1, 0.5, 0.14, 0.07 and 0) by fat content per serving of each item and summing items (Bowes and Pennington, 1994).
Eating behaviors (T3)
At follow-up, eating behaviors related to lowering dietary fat were measured using six items adapted from a validated instrument (Kristal, 1994). Responses were measured on a four-point scale (1 = rarely or never, 4 = always). These questions were asked at follow-up only, to reduce respondent burden and because it had been assumed that the randomized design would minimize baseline differences.
Process measures (T2)
Immediately after program use, intervention group participants were asked questions that assessed how helpful the program was, who it was meant for, how much of the information was new and whether they would be interested in using a similar program in the future.
Intervention
The multimedia tailored intervention was a one-time, 30 min session using a computer kiosk at the Food Stamp office. The program included a 15 min video soap opera, Sisters at Heart, that was written and produced by the project team. The story incorporated nutrition issues into a courtroom drama and love triangle plot. The evidence ultimately shows that the husband died of a heart attack related to poor dietary habits, and the defendant changes to healthier eating habits for herself and her children. The soap opera was designed to entertain and engage participants' interest, model healthy dietary changes through character and plot development, and present information about the risks of unhealthy eating delivered by expert witnesses during the trial and a testimonial at the end of the program.
Interactive `info-mercials' were interspersed in the soap opera to collect survey data from participants and provided individually tailored nutrition feedback based on survey responses. Tailored messages were structured according to the study's conceptual framework (Figure 1
). The project team created a library of tailored audio messages, on-screen graphics and text, and video segments. Algorithms were developed to specify the appropriate feedback messages for each variable or combination of variables from the baseline survey that was used for tailoring. A computer program was then written using Authorware ProfessionalTM software, that used the algorithms to select and assemble the tailored segments into an overall predetermined format. Thousands of different tailored message combinations were possible based on the number of variables used for tailoring, thereby allowing each participant to receive her own customized program.
Participants received immediate feedback to the knowledge questions and were given additional tips to encourage healthy food choices. Baseline food frequency information was used to provide behavioral feedback about participants' dietary fat and was given in graphic form depicting their current fat intake as `high' or `low/moderate'. This evaluation was made by comparing the individual's dietary fat score to a predetermined cutpoint for high fat intake in women (more than 57 g) derived from previous research (Campbell et al., 1994). A specific goal for reducing dietary fat was provided based on which of three food groupings (high-fat meats, high-fat dairy foods or snacks and fried foods) contributed the most fat to the participants' diet. Messages were tailored to stage of change and perceived weight concerns. Pregnant or breastfeeding women received messages that incorporated appropriate dietary advice for calcium and dietary fat intake in the context of balancing health concerns of mother and baby.
After completing the program, intervention group participants were given a StampSmart recipe book providing simple low-fat recipes and a $2 gift coupon for a major grocery store chain. Control group participants received the gift coupon but did not receive any nutrition education or the recipe book. After completing the follow-up survey, control group participants were mailed the recipe book and a nutrition education newsletter. All participants who completed the follow-up survey received a $5 check.
Data analysis
Comparisons between study groups were analyzed at baseline and at follow-up. Because this was a pilot study and measurement procedures differed between study groups at baseline, the analysis of outcome effects was restricted to comparisons between groups at follow-up without adjusting for the baseline values. For continuous measures, F-tests were computed after analysis of variance. For categorical variables, Fisher's exact test (two-tail) and Pearson 
2 tests were used to determine statistical significance.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Descriptive findings
Study participants' mean age was 29.3 years, 32.2% had at least 12 years of education and an average of 2.2 children lived at home. Race/ethnicity was predominantly African-American (85.3%). The study groups were comparable in demographic characteristics at baseline (see Table I
). Most women (72.6%) had high autonomy for food shopping, meal planning and preparation, and 60.1% felt they needed to lose weight. As reflected in Table III, the majority of participants were considering dietary changes to lower fat: 11.5% were in the precontemplation stage, 38.5% in contemplation, 17.7% in preparation and 32.3% in action/maintenance. Dietary fat intake was also related to stage of change, with precontemplators having significantly higher average fat scores (124.7 ± 8.9 g) compared to those in action/maintenance (80.4 ± 5.4 g). A total of 71.4% had fat intake scores above the pre-established cutpoint of 57 g that was used to define `high fat' in the tailored feedback.
|
|
Differences between groups at follow-up
Knowledge
Knowledge of low-fat foods was similar at baseline in both study groups. As shown in Table II
, intervention group members demonstrated significantly higher knowledge scores compared to the control group at follow-up (P < 0.001). In addition, a significantly higher percentage of the intervention group (76.4%) compared to the control group (52.4%) was able to answer all of the knowledge questions correctly.
|
Self-efficacy
Confidence regarding ability to eat low-fat foods was similar between study groups at baseline. A significant increase in self-efficacy (P < 0.001), based on the one-item assessment, was observed in the intervention group compared to the control group immediately after using the computer program, but the difference was not sustained at longer term follow-up (Table II
).
Stages of change
At baseline, significant differences in stages of change were seen between study groups. More individuals in the control group rated themselves in the action/maintenance and precontemplation stages, whereas more intervention group participants were in the contemplation and preparation stages (see Table III). At follow-up, more intervention group participants were in the later stages of change (preparation, action/maintenance) versus the earlier stages (precontemplation, contemplation) compared to controls, 2 (d.f. 1) = 4.58, P = 0.03. In addition, a higher percentage of the intervention group had advanced in stage compared to the control group, 2 (d.f. 1) = 6.25, P = 0.01.
Dietary behavior
Two measures of dietary behavior were used to assess effects of the intervention: the FFQ measured at baseline and follow-up, and the six items related to eating behavior (follow-up only). At baseline the control group reported consuming significantly more fat than the intervention group (P = 0.01). At follow-up, both study groups reported markedly lower mean fat scores compared to baseline levels. Fat scores at follow-up were not significantly different according to intervention condition (Table II). Scores on the eating behavior questions indicated that at follow-up, the intervention group baked meat in the oven significantly more often and consumed low-fat snack items more often than the control group (see Table IV).
|
Process measures
Participants in the intervention group were asked questions to assess their acceptability of the program. The majority of participants (79%) said the program was very helpful and that they would use it again (66%); however, 55% said that none of the information was new. Ninety percent believed the program was for someone like them.
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Results of this pilot study suggest positive effects of a tailored multimedia nutrition education program for low-income women receiving Food Stamps. At follow-up, the intervention group demonstrated higher knowledge of low-fat food choices and a greater percentage had moved forward in stages of change compared to controls. An increase in self-efficacy was seen immediately after using the computer program; however, the difference between study groups was not significant in the follow-up survey data. Both the intervention and control groups reported much lower fat intakes on the FFQ at follow-up and no significant difference was seen between groups. Differences in eating behavior measures suggest that the intervention group was more likely to use low-fat cooking methods and consume low-fat snacks. Process measures indicated high acceptability of the program among this population, who rated it as very helpful and were interested in using a similar program in the future. Participants expressed enthusiasm for the soap opera and one woman remarked that she completed the program to find out `What does murder have to do with nutrition?'.
This study extends the body of literature using computer tailored and multimedia interventions in several areas. First, previous research on the effectiveness of individualized computer-tailored nutrition education has evaluated printed messages as opposed to multimedia programs (Campbell et al., 1994; Brug et al., 1996). These studies targeted mainly European-American and higher socioeconomic status audiences, and were conducted in primary care or worksite settings. The target population for the present study was low-income, primarily African-American women receiving food assistance. The women were pro-actively recruited into the study and were not seeking nutrition education at the time of enrollment. The favorable response to and positive outcomes of the program suggest that tailored multimedia self-help programs may be an effective health promotion strategy to reach underserved populations in settings outside of the health care system, such as social service agencies and community sites. Multimedia may also be more beneficial than print-based media for individuals with limited literacy skills.
Second, the use of formative evaluation methods in designing the program contributed to the level of participation and acceptability that was achieved with a `hard to reach' audience. Personal interviews with Food Stamp recipients identified the expectation of boredom as a major barrier to participation in nutrition education. The subsequent choice to develop a soap opera and interactive tailored `info-mercials' was a novel approach to overcoming that barrier, and suggests that combining interactive education with entertainment preferences of the target group may be an effective way to interest and motivate lower income and minority audiences to participate in health promotion activities. In another pilot study not reported here, effects of the tailored `info-mercials' were assessed with and without the soap opera portion of the program (Honess-Morreale, 1995). The findings showed that knowledge gains were comparable between groups but that the group that did not view that soap opera had lower self-efficacy at the end of the program compared to the group that had viewed the soap opera. This observation is consistent with Social Cognitive Theory, which states that one of the major ways to increase self-efficacy is to view credible role models demonstrating a behavior change (Bandura, 1986). In addition, viewing the soap opera may have enabled participants to use the `dramatic relief' process of change, one of the processes hypothesized to promote progress in the stages of change (Prochaska et al., 1992). Soap operas have been used successfully in interventions targeting issues such as family planning, women's status and adult literacy; however, the authors could find no published studies evaluating their use for nutrition education (Brown and Cody, 1991; Elkamel, 1995).
A third contribution is the finding that program participants demonstrated greater progress in stage of change compared to controls. This finding is encouraging because the program provided behavioral feedback that was tailored to stage. A caution in interpreting these results is that there were significant baseline differences in stage of change between groups, which may have been related to the different modes of survey administration. In addition, despite differences in stage progress, there were no differences in mean fat intake between study groups at follow-up. It is possible that more stage advance in the intervention group did not result in a lower mean fat score due to measurement error and/or the fact that more stage advance in the intervention group made it relatively even with the control group at follow-up in terms of numbers in action/maintenance. The eating behavior questions suggested more behavior change in the intervention group and this measure might have been more responsive to intervention effects than the FFQ (Kristal et al., 1994). Because the eating behavior questions were only asked at follow-up, however, it is not known whether the observed differences were already present at baseline. Future studies should avoid this limitation by including eating behavior questions at both baseline and follow-up.
This study had a number of other limitations. First, it was a one-time, low-intensity intervention. The significant rise in self-efficacy in the intervention group immediately after program use had dissipated at follow-up, suggesting that multiple intervention `boosters' over time might be needed to maintain participants' confidence in ability to lower fat. Using one question to assess self-efficacy, however, may have inadequately measured the construct. Second, the length of time that was necessary to obtain the follow-up data from some participants might have introduced bias, especially if the amount of time to follow-up varied according to study group. Randomization by day, rather than by individual, may also have introduced bias, and indeed the groups were comparable in demographics but showed differences in fat intake and stage. Variables representing time elapsed from baseline to follow-up survey and day of study enrollment were examined but were not found to be predictors of fat intake and stage of change.
Several limitations of the dietary assessment method are noted. First, although formative research was conducted to understand dietary habits, the use of a brief FFQ might have missed important dietary sources of fat in this population. Second, instrumentation bias may have occurred during the study by administering the FFQ using three different methods (computer self-administered, self-administered with research assistant help and telephone administered). The large decreases in reported fat intake observed in both study groups may well have been related to these varying methods of data collection. It appeared that participants' responses to the follow-up telephone-administered FFQ questions provided consistently lower frequency estimates for all items consumed. The lack of observed difference in fat intake between study groups at follow-up, therefore, is difficult to interpret based on dietary data collection methods that may not be comparable. An alternative explanation is that the control group actually changed as much as the intervention group due to a `Hawthorne effect' of simply being in the study. We tend to doubt this explanation based on results of other tailored dietary studies with survey-only control groups, that showed little change in the control group (Campbell et al., 1994; Brug et al., 1996).
Clearly, future studies of computer-interactive health education programs should be designed with sufficient resources and incentives to enable participants to use the same data collection method (i.e. the computer) in both the intervention and control groups at each data collection point. Building on this `lesson learned', public health clients participating in a new study (FoodSmart) are randomized to either intervention or control condition by the computer and complete the baseline survey using the computer. Participants are also scheduled to return to clinic to complete the follow-up survey on the computer, and incentives such as gift coupons and sports bottles are offered. We are finding, however, that a relatively small percentage of participants do return and that telephone surveys may be essential in order to obtain an adequate follow-up response rate in this population.
A final study limitation is the generalizability of the program and results to other audiences. The program was deliberately targeted and tailored to address the interests and preferences of a specific group of low-income women. The resulting program, however, may not be broadly applicable without re-targeting it to each new audience. Optimally, future research should address the need to develop and test tailored programs that contain multiple versions geared to differences in demographics, health issues, language and other characteristics of specific populations and subgroups.
|
Acknowledgments |
|---|
This research was funded by the US Department of Agriculture grant no. 59-3198-3-067. The authors acknowledge additional gifts in support of the project from Food Lion, Inc, and the Quaker Foundation.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Bandura, A. (1986) Social Foundations of Thought and Action: A Social Cognitive Theory. Prentice-Hall, Englewood Cliffs, NJ.
Beresford, S. A., Farmer, E. M., Feingold, L., Graves, K. L., Sumner, S. K. and Baker, R. M. (1992) Evaluation of a self-help dietary intervention in a primary care setting. American Journal of Public Health, 82, 79–84.
Block, G., Hartman, A., Dresser, C., Carroll, M., Gannon, J. and Gardner, L. (1986) A data-based approach to diet questionnaire design and testing. American Journal of Epidemiology, 124, 453–469.
Block, G., Clifford, C., Naughton, M., Henderson, M. and McAdams, M. (1989) A brief dietary screen for high fat intake. Journal of Nutrition Education, 21, 199–207.
Bowes, A, and Pennington, J. (1994) Bowes and Church's Food Values of Portions Commonly Used, 16th edn. Lippincott, Philadelphia, PA.
Brown, W. J. and Cody, M. J. (1991) Effects of a prosocial television soap opera in promoting women's status. Human Communication Research, 18, 114–142.[Web of Science][Medline]
Brug, J., Steenhuis, I., van Assema, P. and De Vries, H. (1996) The impact of a workplace computer-tailored nutrition education instrument. Preventive Medicine, 25, 236–242.[Web of Science][Medline]
Campbell, M., DeVellis, B., Strecher, V., Ammerman, A., DeVellis, R. and Sandler, R. (1994) Improving dietary behavior: the effectiveness of tailored messages in primary care settings. American Journal of Public Health, 84, 783–787.
Campbell, M. K., Symons, M., Polhamus, B., Demark-Wahnefried, W., Bernhardt, J., McClelland, J. W. and Washington, C. (1998) Stages of change and psychosocial correlates of fruit and vegetable consumption among rural African American church members. American Journal of Health Promotion, 12, 185–191.[Web of Science][Medline]
Elkamel, F. (1995) The use of television series in health education. Health Education Research, 10, 225–232.
Honess-Morreale, L. (1995) Effects of a nutrition education program for women. Masters thesis. University of North Carolina, Chapel Hill.
Kristal, A. (1994) Food habits and eating patterns questionnaires. In Thompson, F. and Byers, T. (eds), Dietary Assessment Resource Manual. American Journal of Clinical Nutrition, 124 (Suppl. 11), 2300S.
Kristal, A., Beresford, S. and Lazovich, D. (1994) Assessing change in diet-intervention research. American Journal of Clinical Nutrition, 59 (Suppl. 1), 185S–189S.
Maibach, E. and Holtgrave, D. (1995) Advances in public health communication. Annual Review of Public Health, 16, 219–238.[Web of Science][Medline]
Maibach, E. and Murphy, D. (1995) Self-efficacy in health promotion research and practice: conceptualization and measurement. Health Education Research, 10, 37–50.
McGinnis, J. M. and Foege, W. H. (1993) Actual causes of death in the United States. Journal of the American Medical Association, 270, 2207–2212.
Novelli, W. (1990) Applying social marketing to health promotion programs. In Glanz, K., Lewis, F. and Rimer, B. (eds), Health Behavior and Health Education. Jossey-Bass, San Francisco, CA, pp. 342–370.
Prochaska, J. and DiClemente, 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.[Web of Science][Medline]
Prochaska, J., DiClemente, C. and Norcross, J. (1992) In search of how people change: applications to addictive behaviors. American Psychologist, 47, 1102–1114.[Medline]
Skinner, C., Siegfried, J., Kegler, M. and Strecher, V. (1993) The potential of computers in patient education. Patient Education and Counseling, 22, 27–34.[Web of Science][Medline]
Witte, K. (1995) Fishing for success: using the persuasive health message framework to generate effective campaign messages. In Maibach, E. and Parrott, R. (eds), Designing Health Messages: Approaches from Communication Theory and Public Health Practice. Sage, Thousand Oaks, CA.
Received on February 3, 1997; accepted on January 19, 1998
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  B. Baquero, G. X. Ayala, E. M. Arredondo, N. R. Campbell, D. J. Slymen, L. Gallo, and J. P. Elder Secretos de la Buena Vida: processes of dietary change via a tailored nutrition communication intervention for Latinas Health Educ. Res., April 1, 2009; (2009) cyp022v1. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Thompson, T. Baranowski, K. Cullen, K. Watson, A. Canada, R. Bhatt, Y. Liu, and I. Zakeri Food, Fun and Fitness Internet program for girls: influencing log-on rate Health Educ. Res., April 1, 2008; 23(2): 228 - 237. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Lassen, M. Bruselius-Jensen, H. M. Sommer, A. V. Thorsen, and E. Trolle Factors influencing participation rates and employees' attitudes toward promoting healthy eating at blue-collar worksites Health Educ. Res., October 1, 2007; 22(5): 727 - 736. [Full Text] [PDF]
|
|
|
|
 |
|
 I. Tessaro, S. Rye, L. Parker, K. Trangsrud, C. Mangone, S. McCrone, and N. Leslie Cookin' Up Health: Developing a Nutrition Intervention for a Rural Appalachian Population Health Promot Pract, April 1, 2006; 7(2): 252 - 257. [Abstract] [PDF]
|
|
|
|
 |
|
 J. M. Griffith, J. R. Sorenson, J. M. Bowling, and T. Jennings-Grant Assessment of an Interactive Computer-Based Patient Prenatal Genetic Screening and Testing Education Tool Health Educ Behav, October 1, 2005; 32(5): 613 - 626. [Abstract] [PDF]
|
|
|
|
|
|
A. B. Irvine, D. V. Ary, D. A. Grove, and L. Gilfillan-Morton The effectiveness of an interactive multimedia program to influence eating habits Health Educ. Res., June 1, 2004; 19(3): 290 - 305. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. E. Peterson, G. Sorensen, M. Pearson, J. R. Hebert, B. R. Gottlieb, and M. C. McCormick Design of an intervention addressing multiple levels of influence on dietary and activity patterns of low-income, postpartum women Health Educ. Res., October 1, 2002; 17(5): 531 - 540. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. C. Marcus New Directions for Risk Communication Research: a Discussion With Additional Suggestions J Natl Cancer Inst Monographs, January 1, 1999; 1999(25): 35 - 42. [Abstract] [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||