Blending Sections: When sections blend for a specific purpose
Valeria Gabrielo
Geraldine Garcia
Title? It is compulsory to add titles to your papers.
When the concept of genre comes to discussion, Swales (1990) argues that it is theoretically understood as “a disreputable formulaic way of constructing (…) particular texts” (p.33). However, when practical analysis of genres is carried out by the inspection of sample texts, the idea of mechanical structural construction seems to hold little, if any, validity. The author also states that “communicative purpose has been nominated as the privileged property of a genre. Other properties, such as form, structure and audience expectations operate to identify the extent to which an exemplar is prototypical of a particular genre” (Swales, 1990, p. 52). This theoretical stance can be illustrated by the structural comparative analysis of two research articles, one in the field of medicine and the other in the field of education. A special concern of this analysis will be the demonstration that different sections in the research genre are not canonically organized or discretely presented. On the contrary, they can blend according to the communicative purpose that the author aims at. For this particular study, the result, discussion, conclusion and recommendation sections will be analyzed in an attempt to evidence that even when sections in a text are organized differently, they may have the same purpose.
The result section is characterized by the objective presentation of the resultants of the research carried out. However, considerations pertaining the presentation of the collected data appear to be ultimately determined by the author himself. Pintos and Crimi (2011) explain that the function of the result section is to reveal the outcomes that will either support or refute the original hypothesis of the research. As stated by Swales (1998), the collected data of the research should not be used directly; it should be processed to make the paper feasible to read. In order to summarize the obtained information, the author presents some generalizations further developed by tables, and/or charts (cited in Pintos & Crimi, 2011). In their research paper (Appendix A) Gimbel, Lopes and Nolan Greer (2011) use a result/discussion section in order to present and discuss the data simultaneously. Such section displays the pattern generalization + table + evaluation with accurate resort to past tenses for results' outlining and present tenses for meaning interpretation, thus successfully complying with the requirements of the field. Of utmost significance is the observable coherence between the process-like specification of data collection in the methodology section and the descriptive character of the results/discussion section. However, when it comes to analyzing the use of tables to display information, reading and data interpretation are somewhat obscured. The use of bar charts could have been more suitable to establish comparisons. Besides, the tables show some duplication of the information, which renders them redundant. The sections could be described as unbalanced due to the fact that the amount of information presented outweighs that of the discussion. Different is the example provided by Smith Anderson-Bill's (2011) research paper (Appendix B) where the result section is presented in isolation with significant discussion percolating throughout. In this article, the writer uses a table and figures to complement the information previously introduced in a more effective way. While Gimbel, Lopes and Nolan Greer (2011) follow APA style in the use of tables, Smith Anderson-Bill (2011) does not. Regarding the last sections of the articles under study, Gimbel, Lopes and Nolan Greer (2011) present an implications section which appears to fulfill the function of the conclusion section as it clearly refers back to the hypothesis presented in the introduction. There is also a comparative reference of the obtained results with the information of the reviewed literature. The implications section is followed by a recommendations section which clearly states possible solutions to the problem analyzed throughout the paper with a considerably persuasive effect. Conversely, Smith Anderson-Bill (2011) introduces a discussion section whose function seems to be fulfilled up to a certain extent (Appendix B). While certain elements - namely data interpretation and comments of previous research in the field - are overtly expressed, the ultimate statement of a conclusion is altogether absent from the article. It is precisely this flaw that leaves the reader with an uncomfortable feeling of incompleteness.
Based on the aforementioned analysis, one can aptly conclude that canonical conventions, significant as they may be, are not always decisive in the regular proceedings of the trained scholar. The communicative purpose of the specialized study eventually determines how the written genre will unfold. Different purposes will need to comply with certain pre-established forms, leading the author to make informed choices as to the selection of a specific macro-structure. Likewise, the audience's particular expectations should at no point be overlooked. The present comparative description aimed at demonstrating that, even though authors can approach result, discussion and conclusion sections from different angles arranging them in different ways, there are basic criteria that should not be disregarded. Not only should summarized data be conveniently presented and clearly interpreted, conclusions and/or recommendations for future actions should be provided as well. Only when these aspects have been considered will the research paper achieve its final outcome.
References
Smith Anderson-Bill, E. (2011). Social cognitive determinants of nutrition and physical activity among web-health users enrolling in an online intervention: The influence of social support, self-efficacy, outcome expectations, and self-regulation. Journal of Medical Internet Research, 13. DOI: 10.2196/jmir.1551 http://www.jmir.org/2011/1/e28/
| Papers Analysis (APA 6th ed) |
| Name and Surname:
| Topic:
|
 Draft # Date: | Title: |
| Dimension | Criteria | Points |
| LAYOUT | 1 | 2 | 3 | 4 | (5 to 20) |
|
Format | No headers, no page numbers, no clear margins. Spacing problems. Inappropriate font. | No Header included. Page numbers absence. Spacing problems. | Header and page numbers included. Spacing problems. Inappropriate font. | Clear paper’s presentation. There are page numbers. Respected margins. Correct spacing and type & size of font. 1.5 or double interlining. | 2 |
|
Header |
Not included. | Included. Too much information. Or some info missing. | Included. Not well balanced. | Included. Precise info is given. Well balanced. | 2 |
|
Main Title |
Not included. | Included. Not suitable. Underlined, highlighted or italicized. | Included. Appealing. Underlined, highlighted or italicized. |
Included. Appealing. Centered. Upper & lower cases. | 1 |
|
References |
Not mentioned. Plagiarism. | Mentioned vaguely. Not on a new sheet of paper. Not clear use of references or erroneous sources acknowledgement | Not clear use of references or erroneous sources acknowledgement. | Sources cited clearly in a reference list at the end of the paper. APA style.
| 2 |
|
In-text citations |
Not included. Plagiarism. |
Little use of in-text citations. Incorrect use of required style. |
Included. Not well balanced. Repeated pattern. (e.g. too many quotes, only paraphrasing, etc). |
Included. Well balanced. Different techniques applied. It is read smoothly. APA 6th ed. applied. | 3 |
| CONTENT | 1 | 2 | 3 | 4 | (10 to 40) |
|
Data analysis | Not clear analysis. Relationships & comparisons cannot be followed. Too descriptive. | Brief. Not substantial. Some connections can be followed. Too descriptive. | Clear. Good analysis. No evidence presented. Inversion. Hedging. Conditionals. | Very good. Clear analysis. Comparisons can be established. Evidence is provided. Inversion. Hedging & conditionals. | 3 |
|
Terminology/ Word choice | Difficult to follow. Not understandable. Imprecise language. No acronyms clarification. | Inappropriate terminology. Little clarification. Some terms are not academic. | Legible terminology. Clarification. More academic style. Effective. | Legible terminology. New terms clarification. Effective vocabulary. Good use of connectors. Academic style. | 3 |
|
Spelling | Full of errors. Unreadable. | Many errors. Some parts unreadable. | Few errors. Readable. | All words are spelled correctly. | 4 |
| Sentence variety | Many sentence fragments. Same pattern and length. | Some sentence fragments. Same pattern & length. | Most sentences are complete and varied in pattern & length. | Complete sentences in a variety of patterns and lengths. | 4 |
| Organization
| Vague ideas. Long & confusing intro. Unrelated development. Blurred conclusion.
| Some ideas connected to each other. Purpose established. No transitions. Main point presented. Two of the three parts are not clear or too long.
| Connected ideas. Clear purpose. Marked transitions. One of the three parts is not clear or too long.
| Connected ideas: supporting the main topic. Clear and concise introduction. Clear development: good clarification of major points. Clear conclusion.
| 4 |
| Punctuation
| Frequent and major errors that obscure meaning.
| Some frequent or major errors: Readers’ confusion.
| A few errors.
| No punctuation errors.
| 3 |
| Paragraph content & paragraph length
| Not balanced: too long & too short paragraphs are presented.
| Some paragraphs relate to the topic. Not balanced: too long or too short paragraphs are presented.
| Most paragraphs are related to the topic. Well balanced.
| Paragraph length has been respected & achieved. Smooth. Clear and precise.
| 3 |
| Grammar
| Grammar choices are confusing. Mixture of tenses.
| Some grammar mistakes. Grammar choices sometimes confuse the readers.
| Appropriate grammar choice. No meaning interference.
| Completely appropriate grammar choice: Help readers understand meaning.
| 4 |
| Details
| No or little details (such as explanations, examples, etc) to support & explain the topic.
| Some accurate details. Do not always support topic.
| Accurate info that supports the topic.
| Accurate and relevant info that fully support the topic.
| 4 |
| Tone & audience
| Unclear & inappropriate tone. Audience not considered.
| Inconsistent tone. Incomplete idea of audience.
| Appropriate tone. Audience is considered.
| Appropriate & consistent tone. Audience correctly identified. | 4 |
|
|
|
|
|  Total
| 46/60 |
Comments:
Good job. There are certain things to polish. Edit your paper and upload it to your blog. You make a good team! Your mark is 7 (seven). BTW, I’ve just noticed that the other version of the paper does not have the same cover sheet. Be careful next time. If not, you won’t receive feedback. As the cover sheet was not required, we did not pay attention to this.
Appendix A
Perceptions of the Role of the School Principal in Teacher Professional Growth
Abstract
The purpose of this study was to investigate teacher and principal perceptions of the role of the principal in fostering teachers‘ professional growth. A Likert-type questionnaire was used to explore the ways 476 teachers and 135 principals see themselves as being supported in their professional growth. New and veteran teachers and principals differ in their perceptions of what support they deem important to teacher professional growth. Teachers indicate that having a mentor is the most supportive factor in their growth. Principals tend to agree that listening to teacher concerns is the most supportive factor in fostering teacher professional growth.
Keywords
teacher professional growth; principal-teacher perceptions; teacher development
The value of teacher professional growth, the important role of principals in fostering that growth, and the techniques that are most often used by principals to assist in teacher growth and development have been examined by a number of education scholars in the past (Berube, 2004; Cochran-Smith & Lytle, 1999; Darling-Hammond, 2000, 2005; Drago-Severson, 2007; Dufour, 1995; Glickman, 2002). Most of these studies focus on new and beginning teachers.
What is not clear from the literature is how principals and teachers perceive the behaviors exhibited by principals in promoting the professional growth of teachers.
In this study the researchers examine how principals promote the professional growth of teachers from the perspectives of principals and teachers themselves by describing principals‘ and teachers‘ views on several aspects of principal behaviors.
Currently, there is a national focus on teacher quality. We assert that a contributing factor to teacher effectiveness is how the principal fosters teacher professional growth.
An integral component of sustained school improvement has been the willingness and ability of principals to assume the role as staff developers. To do this, principals must have clear and open communication with teachers and create opportunities to build relationships (Halfacre & Halfacre, 2006;
Youngs & King, 2002). These principal behaviors increase principal-teacher trust, a necessary ingredient in helping teachers reach their professional goals (Gimbel, 2003).
Principal leadership which supports adult development makes schools better places for teaching and learning. Several studies suggest that principals realize that most teachers expand their teaching range only with carefully designed support and assistance (Berube, 2004; Blase & Blase, 1998; Gimbel, 2003; Halfacre & Halfacre, 2006: Sergiovanni, 1992; Zimmerman, 2006).
Findings from these studies point to the principal sharing decision making with teachers and involving them in planning professional development to meet their goals. Teachers tend to demonstrate high self-efficacy when communication with the principal is regular, open and honest (Gimbel, 2003).
Formal and informal opportunities that principals provide for teacher collaboration yield vast positive results for teacher growth. In schools where teachers frequently talk to each other the most about practice and where principals stayed in touch with the community, students had noticeably higher academic achievement (Blase & Blase, 1998; Cochran-Smith & Lytle, 1999; Drago-Severson, 2007; Leanna, 2002; Wenglinsky, 2000).
Results from these studies point to feedback from principals that was particularly helpful for teachers in implementing new ideas, using greater variety in teaching, responding to student diversity, preparing and planning more carefully, taking more risks, achieving better instructional focus, and using professional discretion to make changes.
Findings from studies of narrative feedback written by principals to teachers in their annual evaluations suggest that simply providing general feedback to teachers by the principal did not ―promote and support‖ professional learning.
Methodology
Design
For this descriptive-exploratory study of principal and teacher perspectives, an original questionnaire was used. A list of 20 final questions was developed and critiqued by university colleagues with expertise in questionnaire design. The creation of the final questionnaire emanated from data compiled from a 2-question, field-test questionnaire pilot-tested with a sample of graduate students enrolled in summer graduate courses in education. The 2 questions were:
1. What kind of tangible supports does your principal offer to make you feel you are growing professionally? List 10 behaviors, structures or policies of the principal.
2. What are the barriers to your principal not being able to support your professional growth?
List 10 structures, behaviors, or policies which impede your principal from supporting you professionally.
Method
Following editing, revision, and IRB approval, the final 20-question questionnaire was sent electronically by using Zoomerang, which
guarantees anonymity (Table 2). Teachers are not necessarily rating their own principals. Data were treated and analyzed through the use of SPSS.
Results/Discussion
Demographic data
Respondents included 478 teachers and 135 principals. Elementary principals responded more than those from other grade levels while the greatest number of teacher respondents came from the high school level (Table 1).
Principal respondents were predominantly white females who worked at the high school level for 2-5 years. Teacher participants were predominantly female, white and were likely to work for 2-5 years at the K-5 grade level. In each question, ―n‖ will vary as not all of the 135 principals and 478 teachers responded to each question.
The free/reduced lunch demographic data show that 41.7% of principal respondents came from schools with 5-19% free/reduced lunch while 40.3 % came from the least
Questionnaire
Table 2 results from SPSS show a rank order comparison of behaviors principals and teachers agree most support the professional
growth of teachers. The top 5 ranked by principals are not the same as the top 5 ranked by teachers.
Table 2 shows the percentage of principals and teachers responding to each of the 15 non-demographic questions on the questionnaire. A striking finding from this table is the difference in perception of what teachers, as opposed to principals, indicated as the most important action by the principal that impacts their professional growth. The first ranking supportive action indicated by teachers is ranked eleventh by principals: ―I offer a mentor to new teachers.‖
Further dissonance in perceptions is shown is Table 2. Principals ranked time devoted to listening to teacher concerns as the first supportive behavior for promoting teacher professional growth while teachers ranked the time principals spend listening to them as fourth. Perceptions that teachers have of principals visibly supporting their growth is ranked second by teachers whereas principals rank that action as eighth. The encouragement of teacher collaboration is ranked third by teachers and seventh by principals. Principal visibility is ranked fourth by teachers and third by principals.
A comparison of questions 14 and 15 revealed that twice as many principals (100%) as teachers (45%) responded that they seek teacher input into the decision process.
Our questionnaire responses showed that 94% (Table 2) of principals indicated that they seek teacher input before making a decision and only 45% of teachers reported this is so.
Moreover, 100% of the principal respondents indicated they spent time listening to teachers as an action which influenced teacher professional growth while 78 % of teachers perceived that action as influential to their professional growth.
Ninety seven percent of principals responded that they conducted classroom observations and the same percentage of principal respondents indicated that they offer constructive feedback on instructional practices. Sixty eight percent of teachers reported that their principals conduct observations and evaluations. How can principals report that they support teacher professional growth if only 68% (Table 2) of teachers reported that they felt supported by observation and evaluation and only 56% reported that they felt supported in their professional growth by constructive feedback about their teaching from their principals?
Table 2 shows that about half as many teacher respondents (46%) as principal respondents (95%) reported that the time principals spent speaking informally with them about instructional practice was important to their professional growth. Sixty six percent of teachers indicated that principals acknowledge and recognize their professional growth and 91% of responding principals reported that they do acknowledge such growth.
| Table 2 Rank Order Comparison of Behaviors Principals and Teachers Agree Most Support the Professional Growth of Teachers |
| Principal Questionnaire Item | Principals | Teachers |
| Rank Order | Overall Response | Rank Order | Overall Response |
| I spend time listening to the concerns of my teachers. | 1 | 100% (135/135) | 4 | 78% (368/469) |
| I promote a school climate of open and honest communication among teachers and administrators. | 2 | 99% (134/135) | 9 | 62% (292/475) |
| I am a visible presence to students and teachers. | 3 | 98% (130/132) | 4 | 78% (368/474) |
| I feel comfortable speaking informally with teachers in my school. | 4 | 98% (131/134) | 6 | 76% (363/478) |
| I personally conduct classroom observations and evaluations of teachers. | 5 | 97% (129/133) | 7 | 68% (319/473) |
| I offer constructive feedback on instructional practice. | 6 | 97% (126/130) | 10 | 56% (260/471) |
| I encourage teachers to collaborate. | 7 | 97% (127/132) | 3 | 82% (391/479) |
| I support the professional growth of my teachers. | 8 | 96% (129/134) | 2 | 82% (392/479) |
| I spend time speaking informally with teachers regarding instructional practice. | 9 | 95% (127/134) | 12 | 46% (218/474) |
| I often ask teachers for input before making decisions. | 10 | 94% (126/134) | 13 | 45% (214/473) |
|
|
|
|
|
|
|
|
|
|
| I offer a mentor to new teachers. | 11 | 93% (123/133) | 1 | 84% (398/472) |
| I show recognition and acknowledgement of teachers‘ professional growth. | 12 | 91% (122/135) | 8 | 66% (313/477) |
| I offer an adequate amount of time for teachers to collaborate. | 13 | 69% (92/134) | 11 | 47% (222/474) |
| I procure funds for tuition reimbursement for teachers. | 14 | 62% (82/135) | 14 | 44% (207/474) |
| I have adequate monies to provide professional development for teachers. | 15 | 39% (53/135) | 15 | 30% (139/472) |
A cross-tab analysis of length of employment with the question on recognition and acknowledgement of teachers‘ professional growth (Table 4) shows that 78.3% of first-year teachers reported that principals supported them in this manner and only 59.6% of teachers with more than 21 years experience indicated such recognition by principals.
| Table 4 |
| Question 13: cross- tab analysis by length of employment and " I show recognition and acknowledgement of teachers’ professional growth" |
| Length of employment | Principals n=135 | Teachers n=478 |
| First year | 100% (20/20) | 78.3% (29/37) |
| 2-5 years | 90.1% (46/51) | 72.4% (102/141) |
| 6-10 years | 85.8% (24/28) | 59.8% (73/122) |
| 11-20 years | 85.0% (17/20) | 62.5% (70/112) |
| 21+ years | 93.3% (14/15) | 59.6% (37/62) |
Implications
The purpose for this study was to examine how principals and teachers perceived the role of the principal in facilitating the professional growth of their teachers as determined by self-reported responses of a sample of Massachusetts teachers and principals. The response rate was 8.6% and, as such, this is an exploratory study.
In order to see how similar the 8.6% was to the original sample, we reviewed the composition of the original principal/ administrator sample and saw that the respondent sample paralleled the composition of that sample. For the teacher sample, we had difficulty obtaining email addresses, and therefore, used a purposive sample which reflected the same composition as the principal/administrator sample.
Respondents for both teacher and principal questionnaires reflect similar demographics to the original Massachusetts sample population.
One finding from this exploratory study suggests that the longer teachers are employed, the less the principal seems to recognize their professional growth. If such is the case, this could be demoralizing to veteran teachers, especially those who retool to update their pedagogical and technological skills. The same could apply with regard to principal-teacher communication.
Our data seems to suggest that the longer a teacher‘s tenure, the less communication there is between principal and teacher. This may be a factor in veteran teachers feeling isolated, especially when new teachers arrive at their schools. Further study, with a larger sample and higher response rate may corroborate these preliminary data.
According to our results, teacher respondents do not perceive that principals acknowledged their professional growth, but principal respondents do.
This dissonance in the data may contribute to some teachers feeling unappreciated by their school principals and not being held in esteem for their professionalism. Zimmerman (2006) found that high levels of communication between administration and staff correlated positively with high teacher self-efficacy.
Our literature review demonstrates that strong principal-teacher relationships through both formal and informal evaluations, coupled with ongoing positive dialogue between principals and teachers, are integral to teacher professional growth (Cochran-Smith & Lytle, 1999; Danielson, 2002; Glickman, 2002; Kaplan, 2001; Pancake & Mollier, 2007; Zimmerman, 2006).
Another finding from this exploratory study is the difference in principal and teacher perceptions on the value of having a mentor. For principal respondents, offering a mentor to promote teacher growth does not seem as important as it does to teacher respondents. The first-ranking supportive action indicated by these teacher respondents is ranked eleventh by these principal respondents: ―I offer a mentor to new teachers.‖
Teachers want to feel that their input is valuable in school governance. If they are left out, they feel disenfranchised. Data suggest that principal participants think they seek teacher input before making a decision, but teacher participants do not agree with this perception. Studies conducted by Blase and Blase (1998), Gimbel (2003), and Zimmerman (2006) indicated that teacher input into decision making is important for building principal-teacher trust.
These same authors propose that an open and honest climate is conducive for teacher growth, yet data suggest that such a climate is valued among our principal sample but less so by our teacher sample. Youngs and King (2002), Gimbel (2003), and Zimmerman (2006) suggested that to enhance teacher growth, principals should solicit input from their teachers when making decisions and should maintain open communication with all teachers, new and veteran, to engage them in conversations about instructional practice. In this way, teachers feel validated and respected for their professionalism.
Recommendations
The value of teacher professional growth, the important role of principals in fostering that growth, and the techniques that are most often used by principals to assist in teacher growth and development have been examined by a number of education scholars in the past (Berube, 2004; Cochran-Smith & Lytle, 1999; Darling-Hammond, 2000, 2005; Drago-Severson, 2007; Dufour, 1995; Glickman, 2002).
Three recommendations flow from this exploratory study. First, principals should observe and offer effective, timely feedback to teachers on instructional practice.
Second, the principal‘s role in providing a mentor, especially to new and beginning teachers is important. Teacher data from this exploratory study suggest the importance of a mentor in teacher development.
Principals should look for effective teachers to serve as mentors and provide training for them to serve as role models for their peers. The quality of the teacher mentor, the mentor-protégé relationship, and how the mentor is trained all contribute to the professional growth of the teacher.
Principals need to pay heed to veteran teachers and be sure they are acknowledged for their experience. Additionally, principals need to provide appropriate professional-development opportunities for veteran teachers to grow and contribute to their schools.
Finally, the low response rate may mean that principals and teachers in Massachusetts may be too busy, too disinterested, too distracted, or do not have computer access to participate in an electronic questionnaire. This is disappointing in that the findings may inform practice. Perhaps providing a free course for principals and teachers at our university would increase the sample size. Additionally, the questionnaire could be mailed in a self-addressed, stamped envelope with a follow-up postcard reminder.
Author Biographies
Phyllis Gimbel is associate professor of educational leadership at Bridgewater State College. She is a former secondary school teacher and principal and author of the 2003 book titled
Elizabeth Nolan Greer is an assistant principal at North Smithfield Middle School, in North Smithfield, RI. E-mail: e31fn@yahoo.com The authors would like to acknowledge Pam Russell and Reid Kimball from Bridgewater State University for their contributions to this article.
Appendix B
Social Cognitive Determinants of Nutrition and Physical Activity Among Web-Health Users Enrolling in an Online Intervention: The Influence of Social Support, Self-Efficacy, Outcome Expectations, and Self-Regulation
Eileen Smith Anderson-Bill1, EdD; Richard A Winett1, PhD; Janet R Wojcik2, PhD
Background: The Internet is a trusted source of health information for growing majorities of Web users. The promise of online health interventions will be realized with the development of purely online theory-based programs for Web users that are evaluated for program effectiveness and the application of behavior change theory within the online environment. Little is known, however, about the demographic, behavioral, or psychosocial characteristics of Web-health users who represent potential participants in online health promotion research. Nor do we understand how Web users’ psychosocial characteristics relate to their health behavior—information essential to the development of effective, theory-based online behavior change interventions.
Objective: This study examines the demographic, behavioral, and psychosocial characteristics of Web-health users recruited for an online social cognitive theory (SCT)-based nutrition, physical activity, and weight gain prevention intervention, the Web-based Guide to Health (WB-GTH).
Methods: Directed to the WB-GTH site by advertisements through online social and professional networks and through print and online media, participants were screened, consented, and assessed with demographic, physical activity, psychosocial, and food frequency questionnaires online (taking a total of about 1.25 hours); they also kept a 7-day log of daily steps and minutes walked.
Results: From 4700 visits to the site, 963 Web users consented to enroll in the study: 83% (803) were female, participants’ mean age was 44.4 years (SD 11.03 years), 91% (873) were white, and 61% (589) were college graduates; participants’ median annual household income was approximately US $85,000. Participants’ daily step counts were in the low-active range (mean 6485.78, SD 2352.54) and overall dietary levels were poor (total fat g/day, mean 77.79, SD 41.96; percent kcal from fat, mean 36.51, SD 5.92; fiber g/day, mean 17.74, SD 7.35; and fruit and vegetable servings/day, mean 4.03, SD 2.33). The Web-health users had good self-efficacy and outcome expectations for health behavior change; however, they perceived little social support for making these changes and engaged in few self-regulatory behaviors. Consistent with SCT, theoretical models provided good fit to Web-users’ data (root mean square error of the approximation RMSEA? < .05). Perceived social support and use of self-regulatory behaviors were strong predictors of physical activity and nutrition behavior. Web users’ self-efficacy was also a good predictor of healthier levels of physical activity and dietary fat but not of fiber, fruits, and vegetables. Social support and self-efficacy indirectly predicted behavior through self-regulation, and social support had indirect effects through self-efficacy.
Conclusions: Results suggest Web-health users visiting and ultimately participating in online health interventions may likely be middle-aged, well-educated, upper middle class women whose detrimental health behaviors put them at risk of obesity, heart disease, some cancers, and diabetes. The success of Internet physical activity and nutrition interventions may depend on the extent to which they lead users to develop self-efficacy for behavior change, but perhaps as important, the extent to which these interventions help them garner social-support for making changes. Success of these interventions may also depend on the extent to which they provide a platform for setting goals, planning, tracking, and providing feedback on targeted behaviors. (J Med Internet Res 2011;13(1):e28)
doi:10.2196/jmir.1551
Internet users; dietary habits; physical activity; psychosocial aspects; self-efficacy; social support; self-regulation
A high proportion (83% [1]) of Internet users go to the Web for information on health topics [1-3] including exercise (38% in 2008, up from 21% in 2002) and weight loss (33% in 2008). Although community, health system, and workplace health programs have effectively utilized the Internet for a wide array of behavior-change interventions, the reach of the Internet will be realized through the development of theory-based, purely online interventions for Web-health users [4,5]. Much work remains in developing sound methodology for testing the efficacy of programs delivered online [4]. Despite almost universal Internet access and adoption, researchers know little about Web-health users—the adults who go to the Web to find health behavior and behavior change information and who form the likely participant pool for online health promotion and disease prevention research. Overall, Internet users have been equally either male or female and have tended to be somewhat younger, better educated, and to have higher incomes than the general population [2,3]. Web-health users may be more likely to be female than general Internet users, and those going to the Web for health programs may have poor to fair general health [1]. To our knowledge there have been no studies examining the health behavior and related psychosocial characteristics of potential participants of entirely online health interventions. Generally, attrition in Internet-based health programs is high at 43% to 50% [5], but these figures pertain to participants in programs that use the Internet to deliver programs as part of workplace, primary care, or other community-based interventions. Little is known about how participants interact with stand-alone Web-based health programs, that is, programs that recruit, assess, and intervene entirely online, although early studies have suggested that attrition from such studies may be higher [4]. Similarly, Internet interventions in general tend to recruit many tentative users who attempt but quickly withdraw from programs, fewer short-term users who seem to drop out after using the program for a period, and few stable users who stick with a program over the long-term [6]. With some early evidence that rates of recruitment among Web users making contact with online programs may be low (eg, 8% in a study by Murray et al [4]), it is not clear how adoption or adherence patterns apply or if these patterns are related to participants’ demographic, behavioral, or psychosocial traits. In addition to reflecting potential participants’ characteristics, Web-based health programs should be theory-based and evaluated to validate and refine the application of theory within the Web environment [7-14]. Social cognitive theory (SCT) [15,16] is widely used as the theoretical basis for health behavior change interventions [12] suggesting Internet health interventions must help individuals develop a sense of self-efficacy in specific behaviors (such as being physically active and eating nutritiously), which stems from physically and socially supportive environments and promotes individuals’ positive expectations for behavior change. Higher levels of self-efficacy and expectations of positive outcomes lead to the modification or differential use of self-regulatory skills (ie, planning, self-monitoring, problem solving, self-standards, goals, and self-incentives) essential to maintaining behavior change (see Figure 1 for a schematic representation of SCT). Estimating the initial psychosocial characteristics of users is, therefore, essential to developing effective programs. In previous research, self-efficacy has been associated with healthy nutrition [15,17-21] and physical activity [20,22,23] habits, as has social support from important others, such as family and friends [22,24-26]. Although outcome expectation has been found to contribute beyond self-efficacy to healthy eating habits [17-20], it has not been a consistent predictor of physical activity [27], with some studies suggesting strong support and others revealing a null effect [20,22]. Among people who desire a healthier lifestyle and who have access to healthy foods and infrastructure for physical activity, SCT suggests their success at maintaining behavior change will be determined largely by how well they set goals, plan, and monitor, that is, self-regulate such changes. Outside the obesity and weight-management literatures, self-regulation of nutrition has received scant attention and has often been poorly defined [28]. Nevertheless, self-regulatory behavior has been associated with healthier eating [10,19,22,29-33] and with promoting healthier activity levels in adults [20,22, 34]. The purpose of the present study was to examine the social cognitive determinants of nutrition and physical activity among Web-health users enrolling in a purely online SCT-based nutrition, physical activity, and weight-gain prevention intervention.
| | Figure 1. Social cognitive model of health behavior |
Recruitment and Participants
Web-health users were recruited entirely online for a clinical trial of the Web-based intervention called Guide to Health (WB-GTH) (clinical trials identifier NCT00128570). Advertisements in print and online newspapers in the major media markets of Virginia, Virginia Tech alumni publications, and online solicitations through employer and alumni-related listservs during 3 different time periods created 3 waves of recruitment: September 15, 2007 through January 23, 2008; May 8, 2008 through June 15, 2008; and July 9, 2008 through September 19, 2008. One month of Web-browser ads and 2 local direct mailings were used in wave 1 of the recruitment but yielded very few (ie, < 10) visits to the WB-GTH recruitment website. Print and online newspapers yielded some recruits, but the most effective recruitment strategy was through online alumni and employer publications and listservs. Advertisements and solicitations described the need for participants “18 to 63 years old, residing in the United States or Canada, within our weight guidelines, in good health, and not currently active” for an 18-month research project designed to test an Internet program for improving nutrition and physical activity and prevent weight gain. The Internet program was described as including a walking program “designed for you every step of the way,” a nutrition program “tailored to your needs and preferences,” and a “free pedometer and digital scale.” Preventing weight gain (not weight loss) was emphasized. Potential recruits were informed that involvement in the WB-GTH study would require them to log into the Internet program once a week for 18 months and to complete 3 two-hour assessments. Finally, recruitment materials advised potential participants that in order to be screened for study eligibility they would need to select a user id and password and provide an email address.
Approximately 4700 Internet users visited the WB-GTH site to review project information. About 15% (705) progressed no further than the GTH information page, but during the 3 recruitment waves, 3944 individuals registered for screening: 3024 during the first wave of recruitment, 364 during the second wave, and 556 during the third wave. Registering participants had a mean (SD) age of 42.54 years (12.05 years) and a mean (SD) body mass index of (BMI) of 30.81 (7.32) and were predominantly female (3311 or 84%). Based on self-report, of the 3944 individuals who registered, 88% (3454) were white, 6% (240) were African American; 4% (138) were Asian, and 3% (122) were other. In total, 3% (122/3944) reported Hispanic background.
Eligible Web Users
Of screened Web users, about one-third (1307) met eligibility requirements, that is, they were 18 to 63 years of age (or under 65 at the end of the trial), had high normal to obese BMI (ie, BMI 23 to 39, expanded from BMI 23 to 33 in wave 1, which was deemed unnecessarily stringent), were not currently active (ie, they did not exercise at least 20 minutes 3 times a week), but were otherwise healthy (see Figure 2). The WB-GTH program included a fitness walking component that encouraged participants to gradually move into more vigorous levels of walking exertion; hence, individuals with diagnosed coronary, metabolic or pulmonary disease, or coronary artery disease risk factors as specified by the American College of Sports Medicine [35] were excluded from the sample. Eligible participants had a mean (SD) age of 42.17 (11.17) and were predominantly female (1060 or 81%). Based on self-report, 90% (1177) of the 1307 eligible participants were white, 5% (71) were African American, 2% (21) were Asian, and 3% (38) were other. In total, 3% (34/1307) reported Hispanic background. Of the 1307 eligible Web users, 15% (203) were normal weight (BMI 23 to 24.99), 41% (532) were overweight (BMI 25 to 29.99), 33% (433) were mildly obese (BMI 30 to 34.99), and 11% (139) were obese (BMI 35 to 39.99).
| | Figure 2. Social cognitive model of fiber, fruits and vegetables among web-health users. * P < .05, P < .01, * P < .001 |
Ineligible Web Users
Of Web users screened for the project, two-thirds (2637/3944) did not qualify. A small proportion had overlooked the age requirements listed on the information webpage and were either too old for the research project (n = 24) or declined to provide their ages (n = 3). Almost half of ineligible users did not meet the study’s weight requirements (1206/2637, 46%). The WB-GTH was designed for adults in the high normal to obese weight range so some screened participants were below the weight guidelines (BMI < 23, n = 464), but most who were ineligible were too heavy (n = 742). (As noted above, the BMI cutoff of > 32.9 was modified to BMI ≥ 39 during wave 1 recruitment). A total of 36% (1404/3944) of those who registered were excluded because of medical conditions (n = 922) or because they were too active (n = 482) (see Figure 2 for details). The mean age of ineligible Web users was 42.75 years (SD 12.45 years), similar to eligible users (F1,3942 = 2.40, P = .12), but ineligible recruits were more likely to be female (85% vs 81%, χ21 = 11.88, P = .001) and of nonwhite race/ethnicity (13.6% vs 10%; χ25 = 26.05, P < .001). Although ineligible users were heavier than those who were eligible with a mean (SD) BMI of 31.36 (8.49) versus a mean (SD) BMI of 29.5 (4.13) (F1, 3915 = 62.87, P < .001), the entire range of weights were represented in the ineligible sample, that is, 18% (477) of the 2637 ineligible Web users had a BMI less than 23, 8% (200) had a BMI from 23 to 24.99, 22% (593) had a BMI from 25 to 29.99, 20% (527) had a BMI from 30 to 34.99, 15% (387) had a BMI from 35 to 39.99, and 17% (453) had a BMI ≥ 40. Measures
Participants completed demographic information, physical activity, and psychosocial questionnaires on the WB-GTH website, requiring about 35 minutes. Next, participants were redirected from the WB-GTH site to the NutritionQuest? website where they completed the Block 2005 Food Frequency Questionnaire (FFQ), which required from 30 to 40 minutes. Following each participant’s completion of the FFQ, project staff sent the participant a digital bathroom scale and a pedometer for tracking daily steps taken for 1 week, as described below. Participants were sent 2 email reminders after each assessment component if they did not return to complete the next component within 7 days of the possible completion date. Nutrition
Web-health users completed the Block 2005 FFQ (NutritionQuest?, Berkeley, CA) [36] online. FFQ estimates of intake of daily total fat, percent kcal from fat, daily total fiber, daily fiber grams from beans, daily fiber from fruits and vegetables, daily servings of fruits, daily servings of vegetables, and daily servings of fruits and vegetables combined were examined. Physical Activity
Web-health users used a pedometer (Yamax Digi-walker SW-200, San Antonio, TX) and completed a 7-day walking log provided by the project to record their daily steps taken and their daily minutes walked for exercise. They were to return to the WB-GTH website at the end of 10 days to allow for delivery time and to report at least 4 days of daily steps and minutes walked. The mean (SD) number of days at which participants returned was 15.90 days (6.98 days) excluding 6 participants who began their logs more than 60 days after the logs had been sent. The mean (SD) days of daily steps and minutes walked participants reported at this time was 6.09 days (1.20 days). Mean daily steps and mean daily minutes walked (total steps or total minutes/days recorded) were examined.
Social Cognitive Variables
The Health Beliefs Survey (HBS) [19,20], administered online, measured baseline nutrition- and physical activity-related social support, self-efficacy, outcome expectations, and self-regulation (see Table 1).
| | Table 1. Health Beliefs Survey: Scale descriptions and internal consistency estimates of social cognitive measures |
Statistical Analysis
Latent-variable structural equation modeling (SEM) with LISREL 8.8 (Scientific Software International, Inc, Lincolnwood, IL) [37] assessed the extent to which SCT variables contributed to the nutrition and physical activity behavior of Web users interested in participating in a Web-based nutrition, physical activity, and weight gain prevention intervention. Model fit was evaluated with the Normed Fit Index (NFI) and Nonnormed Fit Index (NNFI) > .90, root mean square error of the approximation (RSMEA) < .05 (P close fit > .05). Chi-square was not used in deference to the large sample size. Latent variables were measured with scores from the FFQ, HBS, and the 7-day walk log. With few exceptions, the distributions of measure scores were skewed or displayed unacceptable kurtosis; measures were normalized using the Blom proportional estimate formula in SPSS version 17.0 (SPSS Inc, Chicago, IL). Additional variables were similarly normalized to retain a consistent unit of measurement within latent variables.
Enrolled Participants
Of 1307 Web users eligible to participate in the WB-GTH baseline assessment phase, 963 (74%) consented to become part of the study. Eligible Web users took an average of about 1 day (mean 1.38 SD 4.51) to enroll and to consent to become part of the study, but this ranged from 1 to 52 days.
Of the 1307 eligible users, 26% (344) either failed to complete consent procedures going no further in the online enrollment process (n = 297) or clicked and confirmed the box “I decline to be part of the study” that was available on all pages of the online consent form (n = 47). Participants who did not consent did not differ in age, racial/ethnic background, gender, or BMI from those who did consent to participate in the study (alpha = .05).
Enrolled Web-health participants had a mean (SD) age of 44.40 years (11.03 years), 83% (803/963) were female, and 91% (873/963) were white. The sample was well educated: participants had completed a mean (SD) of 17.08 (3.3) years of education. Participants also had a median annual household income of about US $85,000, 83% (803/963) were overweight or obese, and 69% (507/735) of those completing the 7-day walk log had step counts in the sedentary to inactive range (ie, < 7500 steps/day). The average (SD) number of steps per day among participants was 6480.31 (2350.86). Most participants lived in the United States, but a small number (42) were Canadian residents. Although 51% (488/963) of participants lived in Virginia, the research location, most states were represented in the study (no participants lived in South Dakota, Louisiana, Rhode Island, or Iowa).
Of the 963 Web users participating, 731 completed all components of the baseline assessment in 11 to 135 days. The average (SD) number of days to completion of the baseline assessment was 22.83 (12.62) days. Although the assessment was designed to be completed across 8 days (1.25 hours online, plus the 7-day walking log), only a small percentage followed the prescribed timeline; 95% (694) completed the assessment within 45 days of enrollment. There were no demographic, social cognitive, or nutritional differences between participants with all assessment components and those without, with one exception. Participants who dropped out of the study prior to completion appeared to have slightly lower self-efficacy for making changes in their nutrition behavior that those who completed. Among those who dropped out during the assessment, the mean (SD) self-efficacy score for avoiding high fat and high sugar foods was 67.83 (22.19) versus 71.97 (19.49) among those who did not drop out (F935,1 = 5.06, P = .03) and the mean (SD) self-efficacy score for tracking nutrition was 79.61 (22.63) among those who dropped out versus 82.87 (17.04) among those who did not (F935,1 = 6.79, P = .009).
Nutrition Characteristics of Web-Health Users
Fat, Fiber, Fruit, and Vegetable Consumption
Table 2 contains the means and standard deviations of Web users’ consumption of fat, fiber, and fruit and vegetable servings. Overall, Web users’ dietary consumption was higher in fat and lower in fruits, vegetables, and dietary fiber than recommended. Most, 56% (494/884), consumed more than the generally recommended 65g of total fat/day, 36% (322/884) reported consuming more than 80g of total fat/day, and almost 20% (172/884) reported consuming more than 100g of total fat/day. Only 13% (115/884) of Web-health users consumed the recommended level of 30% or fewer calories from fat; 78% (690/884) reported getting more than half their calories from fat. Similarly, 13% (115/884) of users met recommended levels of fiber intake (ie, at least 25 g/day); 68% (601/884) reported consuming fewer than 20g of fiber/day. Web-users reported somewhat better levels of fruit and vegetable consumption compared with consumption of fiber and fat with 29% (256/884) of participants consuming the recommended level of at least 5 servings/day and almost half consuming at least 4 servings but the remaining users consuming 3 or fewer servings/day. Nutrition-Related Social Cognitive Characteristics
Participant means and standard deviations on the Food Beliefs Survey section of the HBS are reported in Table 2. Web-health users’ responses to the nutrition social support items suggested that they perceived their family members and friends as being fairly neutral in their support of healthier food choices (ie, scores just under 3 on the 5-point Likert-type scale). Web-health users had positive, but not complete, confidence in their ability to eat healthier foods, avoid high fat and high sugar foods, and keep track of their food choices (ie, scoring 71 to 82 on the 100-point Self-efficacy scale). They seemed to agree that their physical health (eg, weight, blood pressure, and appearance) would improve with healthier food choices (ie, scoring on average 4.3 on a 5-point Positive Physical Outcome Expectations scale). Participants were less concerned (ie, scoring on average approximately 2.9 on each 5-point scale), however, that such changes would result in negative social and self-evaluative outcomes (eg, having less time and energy for others and other activities and dissatisfaction with healthier foods).
| | Table 2. Nutrition and physical activity behavior and social cognitive characteristics of inactive but otherwise healthy adults enrolling in a Web-based health promotion intervention trial |
Finally, Web-health users indicated they had never-to-seldom (rated 1 and 2, respectively, on the Self-regulation scale) planned or tracked healthier food choices in the 3 months before the assessment (eg, keep track of high fat snacks or plan to eat fruit for breakfast). They reported that they occasionally (rated 3 on the scale) did things to reduce fat and sugar and increase healthier food choices (eg, drink water instead of sodas or eat fruit for dessert).
Physical Activity Characteristics of Web-Health Users
Daily Step Counts and Minutes Walked
The Web users in the study were selected based on self-reports of exercising less than 20 minutes 3 times a week in the month preceding the assessment.
Daily Steps
Among the inactive participants, average steps logged over 7 days fell within the low active range [38] (see Table 2); 27% (198/735) of the Web-health users took fewer than 5000 steps/day, 42% (309/735) took 5000 to 7499 steps/day, 24% (228/735) took 7500 to10,000 steps/day, and 8% (56/735) took more than 10,000 steps/day. Daily Minutes Walked for Exercise
Web-health users logged an average of less than a quarter of an hour in daily walking (see Table 2); 41% (299/735) logged virtually no walking (< 3 minutes/day). On the other hand, 22% of the sample logged 20 minutes or more/day in walking (169/735). Physical Activity-Related Social Cognitive Characteristics
Participants’ means and standard deviations from the Physical Activity Beliefs Survey portion of HBS can be found in Table 2. Web users interested in a program to help them become more active generally did not perceive their friends and family members as taking steps to being physically active themselves (ie, social support scores of < 3.0 on the 5-point scale). Physical-activity self-efficacy scores indicated that Web-health users had some confidence in their ability to increase physical activity in the face of social, emotional, and logistical barriers (ie, the mean score was about 65 on a 100-point scale). Within the self-efficacy items on the Physical Activity Beliefs Survey, however, participants’ responses varied. Compared to Web-users’ higher mean (SD) score of 80.02 (19.75) on items regarding managing a walking routine (ie, keeping track of walking, making plans to exercise, and resuming walking after a break), their mean (SD) score of 54.75 (23.52) indicated they were less confident in their abilities to deal with the social aspects of becoming more active (ie, finding someone to walk with, exercising when family wanted more time, or socializing only after meeting exercise goals) (t936 = –40.38, P< .001). Web-health users expected that increasing physical activity would result in health benefits (ie, their mean score was 21 on a 25-point Positive Physical Outcome Expectations scale) and would be good for their mental and physical state (ie, their mean score was 17 on a 25-point Positive Self-evaluative Outcome Expectations scale). Participants were more neutral in their expectations that being more active would interfere with the time they would have for others and other activities (ie, a mean score of 10 on the 25-point Negative Social Outcome Expectations scales).
Overall, Web-health users indicated they had never or seldom (rated 1 and 2 on the scale, respectively) implemented physical activity self-regulation strategies in the 3 months before the assessment (see Table 2). The Web-health users did not track their physical activity (ie, frequency, duration, or intensity of exercise) but were more likely to set goals and plan for being physically active (t936 = 26.66, P< .001) Social Cognitive Determinants of Web Users’ Nutrition and Physical Activity Levels
Nutrition Models
Structural equation analyses evaluated behavioral and social cognitive variables simultaneously to determine how well the SCT models of fat (see Figure 3) and of fiber, fruits, and vegetables (see Figure 4) fit the data collected from the Web-health users. Fit was good for each model; specifically, for the fat model, RMSEA = .045 (95% confidence interval CI? .04 - .05), P (close fit) = .80, NFI = .97, and NNFI = .97. For the fiber, fruit and vegetables model fit indicators were RMSEA = .048 (95% CI .04 - .06), P (close fit) = .66, NFI = .97, and NNFI = .96. The SCT models differed in the amount of variance each explained, which was 14% of fat intake, 22% of fiber intake, and 36% of fruits and vegetables intake. The completely standardized parameter coefficients associated with direct effects of the latent variables in the models are illustrated in Figures 3 and 4. A variable’s direct effect is the portion of its total effect that is independent of other variables in the model; a variable’s indirect effect is the portion of its total effect that is dependent on other variables (covariance matrices and factor loadings associated with the analyses are available from author EA). Social Support and Dietary Intake
Social support from friends and family made a strong contribution (ie, beta total > .20 [39]) to healthier nutrition: Web users who perceived that important others were attempting healthier eating had lower levels of fat (beta total = -.28, P < .001) and higher levels of fiber (beta total = .25, P< .001) and fruits and vegetables (beta total = .34, P< .001). The total effect of social support on Web-health users’ fat intake was largely indirect (beta indirect = -.17, P < .001, indirect/total ratio = .68) through social support’s effect on other model variables influencing fat levels (ie, self-efficacy, beta total = .20, P<.001 and self-regulation, beta total = .67, P < .001). On the other hand, the effect of social support on fiber and fruits and vegetables was entirely indirect (fiber, beta indirect = .34, P< .001, indirect/total ratio = 1.36 and fruits and vegetables, beta indirect = .42, P<.001, indirect/total ratio = 1.23) through self-efficacy (beta total = .17, P < .001) and self-regulation (beta total = .65, P < .001). The large positive indirect effects of social support counteracted small, insignificant negative direct effects on fiber, fruit, and vegetable consumption (see Figure 4).
| | Figure 3. Social cognitive model of fat consumption among Web-health users where * signifies P < .05, signifies P < .01, and * signifies P < .001 |
Self-efficacy and Dietary Intake
Fat intake was also strongly associated with self-efficacy; Web-health users with higher confidence in their ability to make healthier food choices, plan and track food intake, and avoid high fat and high sugar foods reported lower levels of fat on the FFQ (beta total = -.21, P< .001). Self-efficacy did not influence Web users intake of fiber (beta total = .05, P = .27) and fruits and vegetables (beta total = .05, P = .23). Although self-efficacy influenced outcome expectations (negative outcome expectations, beta total = .13, P = .006; positive outcome expectations, beta total = .28, P< .001) and self-regulation (beta total = .16, P< .001) in the fat model, the effect of self-efficacy on fat intake was largely direct (ie, beta indirect = -.02, P = .25; indirect/total ratio = .10).
Outcome Expectations and Dietary Intake
Negative and positive outcome expectations did not exert total effects on the content of Web users’ food intake. This was true for fat (negative outcome expectations, beta total = -.04, P = .37; positive outcome expectations, beta total = .03, P = .47), fiber (negative outcome expectations, beta total = .01, P = .87; positive outcome expectations, beta total = .02, P = .59) and fruits and vegetable (negative outcome expectations, beta total = .02, P = .66; positive outcome expectations, beta total = .03, P = .60). Outcome expectations also did not influence self-regulation as hypothesized by the SCT model (see Figures 3 and 4).
| | Figure 4. Social cognitive model of fiber, fruit, and vegetable consumption among Web-health users where * signifies P < .05, signifies P < .01, and * signifies P < .001 |
