Journal of Pediatric Psychology Advance Access published online on April 29, 2008
Journal of Pediatric Psychology, doi:10.1093/jpepsy/jsn041
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Attributions of Teacher Reactions to Diabetes Self-care Behaviors
1University of Wisconsin-Milwaukee, 2Brown Medical School and Rhode Island Hospital, and 3Medical College of Wisconsin
All correspondence concerning this article should be addressed to Anthony Hains, Department of Educational Psychology, PO Box 413, University of Wisconsin-Milwaukee, Milwaukee, WI 53201, USA. E-mail: aahains{at}uwm.edu
 |
Abstract |
|---|
 Top Abstract MethodsResultsDiscussionReferences |
|---|
Objective This study had two objectives: examine relationships among negative attributions of teacher reactions, anticipated adherence difficulties, teacher support, diabetes stress, and metabolic control; and develop questionnaires to test these relationships. Methods One hundred and two youths with type 1 insulin-dependent diabetes completed instruments measuring attribution of teacher reactions, anticipated adherence, teacher support, and diabetes stress. Metabolic control was measured by percentage of hemoglobin A1c. Results Structural equation modeling demonstrated that negative attributions had direct effects on anticipated adherence difficulties and diabetes stress. Diabetes stress had a direct effect on metabolic control. Negative attributions had a significant indirect effect on metabolic control through associations with diabetes stress. Teacher support moderated the path between negative attributions and anticipated adherence difficulties. Conclusions Youths making negative attributions about teacher's reactions are likely to find adherence difficult in school situations and have increased stress. Results are discussed in terms of a social information processing model of adjustment and practical applications.
Key words: adherence; adolescents; attributions; diabetes; teachers.
Management of type 1 insulin-dependent diabetes (T1D) involves adherence to a series of complex daily behaviors such as monitoring blood glucose levels, injecting insulin, carefully maintaining diet plans, and exercising. Adolescents have more problems with adherence than younger children, despite having greater diabetes knowledge (Anderson, Auslander, Jung, Miller, & Santiago, 1990
; Johnson, Silverstein, Rosenbloom, Carter, & Cunningham, 1986; Thomas, Peterson, & Goldstein, 1997). Social situations may negatively impact the adherence behavior of youths (Wysocki, Greco, & Buckloh, 2003), and youths report that adherence difficulties are more frequent in social and peer contexts (Berlin et al., 2006).
Youths with T1D who make negative attributions about expected friend reactions to their diabetes care efforts are more likely to anticipate adherence difficulties (Hains, Berlin, Davies, Parton, & Alemzadeh, 2006; Hains et al., 2007). They also become increasingly influenced by expected disapproval from friends in situations that require adherence behavior and consequently report less regimen adherent solutions to these problem situations (Thomas et al., 1997). A similar pattern of results is found when youths considered self-care behavior around peers other than friends in school settings (Hains et al., 2007). Furthermore, negative attributions of friend and other peer reactions, and anticipated adherence difficulties have significant indirect effects on metabolic control through associations with diabetes-related stress (that is, the degree of everyday stressors related to living and coping with T1D; Hains et al., 2007).
Friend support does not seem to have a large impact on the relationship between expected friend and peer sanctions for public self-care behavior and anticipated adherence difficulties (Hains et al., 2007). Even with support of friends readily available (Helgeson, Reynolds, Shestak, & Wei, 2006; La Greca, Bearman, & Moore, 2002), youths with diabetes may still worry about friend or peer reactions. Thus, youths’ problems with adherence may be related in part to their own inaccurate thoughts and beliefs.
A social information processing model of adjustment has been proposed as a framework for understanding the role of cognitive appraisals involved with diabetes care efforts around others, and their relationship with diabetes stress and metabolic control (Hains et al., 2006). These appraisals include filtering only specific aspects of the situation, incorrectly appraising others’ intentions, or assessing ambiguous situations as threatening in terms of potential consequences (Crick & Dodge, 1994). The model suggests that youths with T1D may perceive adherence behavior as difficult in social situations with their friends due to a fear of negative evaluations, even when friends are actually supportive. Therefore, maladaptive interpretations of events could result in poor behavioral adherence choices and/or emotional distress, and potential problems with metabolic control. In order to more fully understand the role of negative attributions concerning the reaction of others to self-care efforts, it is important to extend the model to examine other important people in the lives of these youths, such as teachers.
While adolescent youths spend a significant amount of time with friends and other peers, they also spend a major part of their day during the academic year under the supervision of teachers. The support and care provided by teachers likely play an important role in successful self-care and health (Wagner, Heapy, James, & Abbott, 2006). Unfortunately, parents and students report concerns about the diabetes knowledge of school personnel and inflexible approaches on the part of schools to help with regimen adherence in the classroom (Nabors, Lehmkuhl, Christos, & Andreone, 2003).
Although decisions to avoid adherence behavior have been found to be related to negative attributions of friends and peers, previous research has not examined the potential relationship between negative attributions of teacher reactions to self-care efforts in school settings and perceived difficulties in adherence. Likewise, no information is available on whether negative attributions of teacher reactions or adherence difficulties is related to diabetes-related stress and metabolic control. Since youths spend a great deal of time around teachers, it is important to find out what youths perceive and how they feel about school situations requiring public displays of adherence. A relationship between negative attributions, adherence difficulties, diabetes stress, and metabolic control might be lessened within the context of high levels of teacher support, unlike the support of friends or peers. Teachers, being adults, may be seen as more reliable, predictable, or mature than same age peers. In addition, youths are likely to have multiple teachers on a daily basis, who display varying levels of support. Teachers perceived as more or less supportive by youths may further impact the nature of this relationship.
This study, which was part of a larger research project investigating youth attributions concerning self-care efforts in social situations (Hains et al., 2007), was designed to (a) clarify the relationships among negative attributions of teacher reactions, anticipated adherence difficulties, most supportive and least supportive teacher support, diabetes stress, and metabolic control; and (b) develop questionnaires with sufficient psychometric properties to test this aforementioned model. The hypothesized model is presented in Fig. 1. Using structural equation modeling (SEM), the first objective was to examine whether a similar pattern of relationships held for teachers that holds for friends and peers. That is, we assessed whether there was an indirect relationship between negative attributions of teacher reactions and metabolic control through the former's influence on anticipated adherence difficulties and diabetes stress. Second, we examined whether the support of teachers identified as most and least supportive by participants differentially moderated the relationships among attributions, adherence difficulties, stress, and metabolic control. We asked the participants to identify most and least supportive teachers (e.g., the two extremes) because youths likely experience a range of teacher behavior and support styles. Understanding of the range of influence and specifics of teacher support (as opposed to overall school environment and average support) would likely provide useful, contextually specific, and nuanced information. Lastly, we hypothesize the Teacher Attribution Questionnaire would demonstrate a good model fit using confirmatory factor analysis and acceptable reliability and validity.
|
|
Methods |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
Participants
The study was reviewed and approved by the institutional review boards of the University of Wisconsin-Milwaukee and Children's Hospital of Wisconsin (CHW). English-speaking youths between the ages of 10 and 18, who were outpatients receiving treatment for T1D in the Diabetes Clinic at CHW, were recruited for the study. The participants are seen for management and treatment of their diabetes in outpatient appointments at the clinic every 3–4 months. Written informed consent/assent was obtained from a parent/guardian and from the youth. Initially, 123 youths consented to participate and 102 (83%) returned the instruments. The mean age of the participants was 13.87 years (SD = 2.01, range 10–18 years) and 60% were females. The average time since diagnosis was 5.58 years (SD = 4.1 years) and ranged from 3 months to 16 years. The self-reported racial background of the final sample included 81 European Americans, six African Americans, four Latino/Hispanics, two Asian Americans, two American Indians, and five multiracial individuals.
Procedure
Participants were part of a larger study examining the role of attributions of others’ reactions to their adherence behavior in public settings (Hains et al., 2007). Data on metabolic control and diabetes-related stress have been previously published (Hains et al., 2007). Participants were recruited by one of two methods. In one method, members of the diabetes treatment team introduced the research study to youths and their parents while they were at a clinic visit. If the families expressed interest, a graduate student in psychology described the project in more detail and obtained consent. In a separate method of recruitment, the study was introduced to the youths and their families at an evening educational group hosted by the diabetes clinic at CHW. Graduate students in psychology staffed an information booth during the evening's activities. A clinic team member hosting the evening session introduced the graduate students and directed interested individuals to this information booth. The graduate students provided a description of the study to interested parties and obtained consent.
Participants were given a packet of instruments that included demographic information, attributions about reactions of others to self-care in social situations, support, and diabetes stress. Youths were given the option of completing the instruments at clinic or the evening session, or taking them home and returning them by paid business reply envelope. Any participant not returning the forms within 2 weeks was given a phone call. The majority of the youths who returned completed questionnaires did so within 2 weeks. All youths completing the questionnaires received a $20 gift certificate to a local shopping mall for their participation.
Measures
Demographic Information
Demographic information related to gender, race, age, grade in school, and duration of diabetes was included on a cover sheet.
Attribution of Teacher Reactions
The Teacher Attribution Questionnaire was adapted from an earlier version of an attribution questionnaire, which focused only on friend reactions (Friend Attribution Questionnaire; Hains et al., 2006). This previous measure has demonstrated good reliability (internal consistency) and concurrent, criterion, and construct validities (Hains et al., 2006). The current questionnaire described five school situations involving teachers where the youth is faced with an adherence situation. Specific situations were developed based on the clinical experience of the researchers.
The following is an example of an adherence situation involving teachers: "You are in gym class and you are playing soccer. You start to feel light headed and shaky after 15 minutes. The gym teacher is starting to criticize your play." Some of the situations specified certain teachers (e.g., gym class or math class). Others involved interactions with unspecified teachers. Each of the five teacher situations was followed by 12 questions asking the youths how they think their teachers would react if they engaged in self-care in the situation.
Several of these questions asked to what extent the youths expected to have certain thoughts about the teachers’ reactions, with the youths responding on a 5-point scale (1 = "Strongly Disagree" to 5 = "Strongly Agree"). Positively worded attributions (e.g., I'd think my teachers would understand) were reverse scored. Five of these questions were averaged to form the Negative Attribution of Teacher Reactions (NATR) scale. One additional attribution question was not included in the scale to parallel the analyses conducted on the Friend Attribution Questionnaire from which it was adapted. These previous factor analyses dropped the item because it did not load. Five additional questions asked about the ease and likelihood of adherence in these situations, which were summed and averaged to form the Anticipated Adherence Difficulties–Teachers (AAD–T) scale. One additional question asked youths whether they had been in experiences like the one presented in the scenario, in order to assess the ecological validity of the vignettes and did not factor into the sub-scales. Both scales were coded to reflect increasing levels of the named construct (e.g., high AAD–T scores reflect more reported levels of anticipated adherence difficulties around teachers).
Diabetes Stress
The Diabetes Stress Questionnaire (DSQ; Boardway, Delamater, Tomakowsky, & Gutai, 1993) is a 65-item self-report instrument designed to assess daily stressors for youths related to diabetes. The measure yields a composite scaled score with higher scores indicating higher levels of stress. Internal consistency has been reported to be excellent (Cronbach's 
=.97), and the measure has also been shown to have good concurrent validity (Boardway et al., 1993).
Teacher Support
Teacher support was examined with the Diabetes Teacher Support Questionnaire (DTSQ). The DTSQ was adapted from the Diabetes Social Support Questionnaire (DSSQ; Bearman & LaGreca, 2002), a 28-item self-report measure of friends’ support for diabetes care. Internal consistency for the DSSQ is high (Cronbach's >.90), and the measure has been found to have good correspondence with other support measures (Bearman & LaGreca, 2002). Items were selected from the DSSQ based on their relevance to the measurement of teacher support. The DTSQ asked youths to rate the frequency with which a teacher engages in a list of 11 diabetes-related support behaviors (e.g., "encourage exercise") and their corresponding feelings about each of these 11 behaviors. Frequency items were rated on a 6-point scale (0 = "never" to 5 = "at least once a day"). Feelings items were rated on a 5-point scale (–1 = "not supportive" to 3 = "very supportive"). Participants provided frequency and feelings ratings for each of these 11 items for a teacher they viewed as most supportive and then again for a teacher they viewed as least supportive, with respect to diabetes. Thus, support scores were available for both most supportive teachers (MSTs) and least supportive teachers (LSTs). For the purposes of this study only the frequency scores for the most and LSTs were used.
Metabolic Control
Metabolic control of the sample was measured by the percentage of hemoglobin A1c (HbA1c). HbA1c was obtained from the clinic visit during which the youths were recruited or the most recent clinic visit in the case of participants recruited during the evening educational group. All samples were collected via DCA2000 (Bayer, Tarrytown, NY, USA) with the nondiabetic reference range between 4.5% and 5.7%. HbA1c levels reflect the average level of blood glucose over a 2- to 3-month period. The mean HbA1c level for the participating youths was 8.31 (SD = 1.38), which is comparable to the mean for the clinic as a whole (M = 8.6).
|
Results |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
Analytic Plan
In order to test the proposed model a series of steps was needed: (a) determine adequacy of the factor structure and psychometric properties of the Teacher Attribution Questionnaire using confirmatory factor analysis (CFA); (b) test the hypothesized relationships among the study variables; (c) determine significance of the indirect effects; and (d) determine whether teacher support moderates relationships within the models. For all analyses, list-wise deletion was used.
To assess the fit of the measurement and structural models, a variety of indices appropriate for smaller samples and nonnormal data were used including a Satorra–Bentler Scaled chi-square (SB 
2) to degrees of freedom ratio of two or less (Ullman, 2001), a comparative fit index (CFI) above.90, a root mean square error of approximation (RMSEA) statistic below.10, and a standardized root mean square residual (RMR) below.08 (Browne & Cudeck, 1993; Hu & Bentler, 1999; MacCallum, Browne, & Sugawara, 1996).
To obtain initial evidence for the validity of the measures, the magnitude and direction of the correlations among study variables were examined. Internal consistencies were calculated as an index of reliability.
Youths who have had T1D <1 year are frequently excluded from studies due to being in the "honeymoon" period. Preliminary analyses were conducted to determine if deleting the 10 participants in this study with <12 months since diagnosis (M = 5.7 months, SD = 2.98, range 3–11 months) appreciably changed the relationships among study variables. The strength and p-values of these relationships were essentially the same in all but one instance. The one exception was the correlation between diabetes stress and the frequency of the MST's support, which was modestly higher in the sub-sample of children with T1D >1 year (r =.21, p =.05) compared to the total sample (r =.16, p =.13). It was decided to retain all participants in the subsequent analyses for three reasons: (a) teacher support was hypothesized as a moderator and as such its direct effect on diabetes stress was not of interest; (b) the difference in this relationship across sub-samples was not statistically significant (z = 1.53, p =.13); and (c) to increase statistical power and stability of parameter estimates with a largest possible sample.
Factor Structure and Psychometric Properties of the Teacher Attribution Questionnaire
CFA using a robust maximum-likelihood estimation method in LISREL 8.54 (Jöreskog & Sörbom, 2003) was employed to validate the factor structure of the Teacher Attribution Questionnaire (TAQ). This method allowed for the computation of a SB 2 and robust standard errors, which adjust for multivariate kurtosis (Satorra & Bentler, 1994). Similar to the peer and friend attribution questionnaires (Hains et al., 2006), the hypothesized model allowed the error term between the frustration and upset items to be correlated. These initial analyses indicated that the proposed models provided an excellent fit to the data according to Bollen's (1989) and Hoyle's (1995) standards: SB 2/df = 1.74 (57.36/33), NNFI =.95, CFI =.96, RMSEA = 0.086, SRMR = 0.069 and provided evidence for content validity.
With regard to the validity of the measures, substantiation of criterion validity was obtained by examining the magnitude and direction of the correlations among study variables (Table I). More specifically, NATR had significant and large to medium positive relationships with diabetes-related distress and AAD–T and no direct relationship with metabolic control (Table I). Similarly, AAD–T was positively correlated with a medium-sized relationship with diabetes-related distress and not related directly to metabolic control. These results provide evidence of criterion validity, since these relationships are in the magnitude and direction that are theoretically expected (Cohen & Swerdlik, 1999). Furthermore, the corrected item-total correlations for the subscales of TAQ (NATR: M = 0.56, SD = 0.11; AAD–T: M = 0.59, SD = 0.08) provided strong evidence for both content and construct validity (Cohen & Swerdlik, 1999). With regard to reliability, the internal consistency coefficients were excellent for the NATR and AAD-T scales (Table II). Table II shows the means, standard deviations and internal consistencies () for the various scales and scale items across vignettes used in this study.
|
|
Hypothesis Testing
Model Evaluation
In light of the preliminary evidence for the factor structure and reliability of the TAQ, the hypothesized model was evaluated using SEM. Although similar to multiple regression (MR), there are several benefits to testing models via SEM. These benefits include the ability to: (a) simultaneously estimate the significance of both direct and indirect relationships; (b) obtain estimates corrected for variables with nonnormal distributions and less than perfect reliability; and (c) determine the adequacy of the model using goodness-of-fit statistics. Compared to testing mediation MR (where prerequisite conditions must often be met), SEM is a more flexible analytic strategy, allowing researchers to test models with hypothesized indirect effects (where a predictor and criterion relationship is not required) and/or mediated effects (see Holmbeck, 1997
for additional details on the distinction between indirect vs. mediated effects in MR and SEM). With SEM, however, great care must be taken to not infer causation or a longitudinal nature with cross-sectional data (Maxwell & Cole, 2007). To decrease the number of estimated parameters, latent variables were defined by fixing the sole indicator's factor loading to 1.0 and its error term to 1 minus the reliability (-coefficient) multiplied by the indicator's variance (Hayduk, 1987). This technique produces path estimates identical to those obtained using multiple indicators or item parcels (Sass & Smith, 2006). This technique allowed for this study's cases to measure ratio (approximately 14:1) to exceed standard recommendations for SEM of five to ten cases per measure (Bentler, 1990; Kline, 2004). For HbA1c, the error term was set to 10% given that the correlations of approximately 0.95 between the DCA2000 and high-performance liquid chromatography, suggest reliable variance of around 90% (Guerci et al., 1997).
The hypothesized model was tested using the robust maximum-likelihood estimation method in LISREL 8.54. All goodness-of-fit indices suggested an excellent fit between the models and the data: SB 2 (df = 2) = 0.79 (p = 0.67); RMSEA = 0.00 (90% = 0.0; 0.16); CFI = 1.00; Standardized RMR = 0.029. As predicted, negative attributions had a direct effect on anticipated adherence difficulties, and diabetes stress had a direct effect on metabolic control. Contrary to the hypothesized model, AAD–T was not significantly related to diabetes stress when controlling for the effect of negative attributions.
As hypothesized, a significant indirect relationship between negative attributions of teachers’ reactions and metabolic control was found through the intermediary variable of diabetes stress. This model accounted for the following percentages of variance of dependent latent variables: 54% AAD–T, 21% diabetes stress, and 6% metabolic control. This final model with standardized estimates (for ease of interpretation and comparison) is presented in Fig. 2 with the indirect relationships indicated with a dashed line. Additional details regarding the model, the covariance/correlation matrixes of items and variables, and unstandardized estimates can be obtained from the corresponding author.
|
Moderation Analyses
When MRs are used to test moderations it can produce biased and inconsistent coefficient estimates along with a loss of statistical power as the reliability of the measures decline (Aiken & West, 1991
; Busemeyer & Jones, 1983). To correct for reliability, latent variable factor scores were created from a measurement model where the latent simple effect variables were defined in the manner previously described. Interaction variables were created by taking the products of the (error-free) latent variables. These latent variables were used with MR to determine whether the most and least supportive teacher's support moderated the significant paths between (a) NATR and anticipated adherence difficulties among teachers; (b) NATR and diabetes stress; (c) diabetes stress and metabolic control. The results of these MRs can be found in Tables III and IV for the MST and LST respectively, with the results of post hoc probing (simple slopes and t-values) of significant interactions at –1 SDs, the mean, and +1 SDs of the moderator presented in Table V.
|
|
|
With regard to specific analyses, the frequency of the MST's support was found to moderate the path from NATR to AAD–T (Table III), with follow-up analyses revealing that MST's support buffered (lessened) these aforementioned paths (Table V). The frequency of the LST's support was also found to significantly moderate the path from NATR to AAD–T, however in contrast to MST, the relationship between NATR and AAD-T increased at higher levels of the LST's support (Table V). A trend toward significance (p =.06, Table IV) was also found with increasing levels of LST's support decreasing the path from NATR to diabetes stress (Table V).
|
Discussion |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
This study was designed to clarify the relationships among negative attributions of teacher reactions, anticipated adherence difficulties, most supportive and least supportive teacher support, diabetes stress, and metabolic control. Negative attributions of teacher reactions had a direct effect on anticipated adherence difficulties and a direct effect on diabetes stress. Diabetes stress, in turn, had a direct effect on metabolic control. In addition, negative attributions of teacher reactions had a significant indirect effect on metabolic control through associations with diabetes-related stress. The present findings parallel past research indicating that negative attributions of friends’ and peers’ reactions, and anticipated adherence difficulties had significant indirect effects on metabolic control through associations with diabetes-related stress (Hains et al., 2007
). In the current study, however, negative attributions of teacher reactions had a stronger association with diabetes stress than adherence difficulties. In addition, only diabetes stress had a direct effect on metabolic control. The role of teacher support (for both most and least supportive teachers) was found to moderate the path between negative attributions and anticipated adherence difficulties. In the case of teachers identified as being most supportive, increased teacher support may lessen the impact of attributions on anticipated adherence difficulties. Negative expectations of teacher's reactions become less important if the MSTs are engaging the youths in desired ways. On the other hand, increased support from LSTs may exacerbate the relationship between negative attributions and adherence problems. Under these situations, the youths may be less appreciative of the efforts or find the support intrusive.
A secondary aim of this study was to develop a questionnaire that could assess appraisals about teacher reactions to diabetes self-care and anticipated adherence difficulties around teachers. This article provided some preliminary evidence for validity and reliability for this measure. This measure may be a helpful tool for researchers or clinicians interested in assessing these domains in youths with diabetes.
There are a number of limitations to this study. First, youths who have negative expectations of teacher reactions may actually have had frustrating experiences performing diabetes care in school. Indeed, youths have reported concerns with lack of flexibility in school personnel when it comes to adherence (Nabors et al., 2003). In addition, participant responses to these school vignettes may not provide an accurate account of actual behavior in real situations with teachers. Also, this study did not take into account the youths’ preference for the type and style of support received. For example, some youths may appreciate a supportive style that is more involved, whereas others may prefer teachers are minimally involved or who are involved only at the request of the youth (Sato et al., in press).
There are a number of methodological concerns as well. For example, data were not coded to assess differences between the two recruitment methods. Likewise, data on demographics and metabolic control were missing for those youths who did not return completed forms and therefore we cannot compare them to youths who completed the study. Information on the type of diabetes regimen (e.g., insulin pump vs. injection) used by the participants was not collected. Various regimens likely have differing opportunities to perform diabetes management at school, which might impact expectations or experiences. In terms of the sample, the majority of youths completed the questionnaires at home, and we have no information on whether they completed the forms independently. While most youths had been diagnosed over a year, a few youths with relatively recent diagnoses were included, during a time when the "honeymoon period" may be playing a major role in metabolic control. Also, the sample was predominately European American (81%), which limits the generalizability of the findings. Because of the cross-sectional design of the study and the use of predominantly self-report measures from a single respondent, causal relationships cannot be specified and the strength of relationship may be due in part to shared method variance. Finally, the current study may not have had enough statistical power to detect small to medium moderating effects of teacher support (e.g., the path between negative attributions and anticipated adherence difficulties among teachers) or any three-way interactions involving gender (Aiken & West, 1991).
Results of the study point to the role of negative appraisals of teacher reactions to self-care efforts in school situations. Similar to their interactions with friends and peers in social situations, youths who expect social sanctions from teachers for engaging in self-care behavior at school anticipate more adherence difficulties. Unlike the role of friend support, however, teacher support may have an impact on this relationship. That is, while the social information processing model of adjustment suggests that youths’ problems with adherence may be related in part to their own inaccurate thoughts and beliefs when it comes to dealing with teachers, this relationship is moderated by whether the teacher is seen as more or less supportive. Offers of support by trusted and well-liked teachers may minimize the impact of negative attributions on behavior. Desired behavior by teachers seen as supportive may serve as a protective factor against adherence problems. Youths may perceive these adults as a source of stability in their daily lives, whereas peer support may be seen as fleeting or inconsistent. Alternatively, offers of support by teachers seen as less supportive may increase the association between negative attributions and adherence behavior. Under these circumstances, the youths may not see the support as genuine.
There are a number of practical applications that are indicated by this research. First, practitioners and school counselors can readily work with youths to identify and approach those teachers perceived as supportive and enlist their help during the school day. This support may be as simple as seeking an advocate for a troublesome issue related to self-care in school. Second, many teachers are naturally interested in the health and well-being of their students. Many school personnel, however, lack specific knowledge related to diabetes care. Diabetes education programs for teachers need to be implemented, in order to enhance support for adherence in school (e.g., recognizing signs of hypoglycemia, making accommodations for students to test blood sugar) (Siminerio & Koerbel, 2000). Finally, cognitive behavioral interventions focusing on restructuring potentially negative attributions of teacher reactions related to self-care are indicated by the results of this study. In addition, behavioral strategies to improve access to teacher support (e.g., assertiveness skills to explain self-care needs in generally structured school situations that typically would not permit such behavior, such as snacking in class), or problem-solving around issues related to teachers perceived to be unsupportive could be included in a treatment package.
Conflicts of interest: None declared.
Received October 22, 2007; revision received March 31, 2008; accepted April 4, 2008
|
References |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
Aiken LS, West SG. Multiple regression: Testing and interpreting interactions (1991) Newbury Park: Sage Publications.
Anderson BJ, Auslander WF, Jung KC, Miller JP, Santiago JV. Assessing family sharing of diabetes responsibilities. Journal of Pediatric Psychology (1990) 15:477–492.
Bearman KJ, LaGreca AM. Assessing friend support of adolescents’ diabetes care: The Diabetes Social Support Questionnaire-Friends Version. Journal of Pediatric Psychology (2002) 27:417–428.
Bentler PM. Comparative fit indexes in structural models. Psychological Bulletin (1990) 107:238–246.[CrossRef][Web of Science][Medline]
Berlin KS, Davies WH, Jastrowski KE, Hains AA, Parton EA, Alemzadeh R. Contextual assessment of problematic situations identified by adolescents using insulin pumps and their parents. Families, Systems, & Health (2006) 24:33–44.[CrossRef]
Boardway RH, Delamater AM, Tomakowsky J, Gutai JP. Stress management training for adolescents with diabetes. Journal of Pediatric Psychology (1993) 18:29–45.
Bollen KA. Structural equations with latent variables. (1989) Oxford, England: John Wiley and Sons. 1989.
Browne MW, Cudeck R. Alternative ways of assessing model fit. In: Testing structural equation models—Bollen KA, Long JS, eds. (1993) Newbury Park, CA: Sage. 136–61.
Busemeyer JR, Jones LE. Analysis of multiplicative combination rules when the causal variables are measured with error. Psychological Bulletin (1983) 93:549–562.[CrossRef][Web of Science]
Cohen RJ, Swerdlik ME. Psychological testing and assessment: An introduction to tests and measurement (1999) 4th. Mountain View, CA: Mayfield Publishing Company.
Crick NR, & Dodge K. A review and reformulation of social information-processing mechanisms in children's; social adjustment. Psychological Bulletin (1994) 115:74–101.[CrossRef][Web of Science]
Guerci B, Durain D, LeBlanc H, Rouland JC, Passa P, Godeau T, et al. Multicentre evaluation of the DCA 2000 system for measuring glycated haemoglobin. Diabetes & Metabolism (1997) 23:195–201.[Web of Science][Medline]
Hains AA, Berlin KS, Davies WH, Smothers MK, Sato AF, Alemzadeh R. Attributions of adolescents with type 1 diabetes related to performing self-care around friends and peers: The moderating role of friend support. Journal of Pediatric Psychology (2007) 32:561–570.
Hains AA, Berlin KS, Davies WH, Parton EA, Alemzadeh R. Attributions of adolescents with type 1 diabetes in social situations: Relationship with expected adherence, diabetes stress, and metabolic control. Diabetes Care (2006) 29:818–822.
Hayduk LA. Structural equation modeling with LISREL (1987) Baltimore, MD: Johns Hopkins University Press.
Helgeson VS, Reynolds KA, Shestak A, Wei S. Friendships of adolescents with and without diabetes. Journal of Pediatric Psychology (2006) 31:194–199.
Hoyle RH. Structural equation modeling: Concepts, issues, and applications (1995) Thousand Oaks, CA: Sage.
Holmbeck GN. Toward terminological, conceptual, and statistical clarity in the study of mediators and moderators: Examples from the child-clinical and pediatric psychology literatures. Journal of Consulting and Clinical Psychology (1997) 65:599–610.[CrossRef][Web of Science][Medline]
Hu L, Bentler PM. Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling (1999) 6:1–55.
Johnson SB, Silverstein J, Rosenbloom A, Carter R, Cunningham W. Assessing daily management in childhood diabetes. Health Psychology (1986) 5:545–564.[CrossRef][Web of Science][Medline]
Jöreskog KG. LISREL 8: User's reference guide (2003) Inc; 2000: Lincolnwood, IL: Scientific Software International.
Kline RB. Principles and practice of structural equation modeling (2004) 2nd. New York: Guilford.
La Greca AM, Bearman KJ, Moore H. Peer relations of youths with pediatric conditions and health risks: Promoting social support and healthy lifestyles. Developmental and Behavioral Pediatrics (2002) 23:271–280.
MacCallum RC, Browne MW, Sugawara HM. Power analysis and determination of sample size for covariance structure modeling. Psychological Methods (1996) 1:130–149.[CrossRef][Web of Science]
Maxwell DE, Cole DA. Bias in cross-sectional analyses of longitudinal mediation. Psychological Methods (2007) 12:23–44.[CrossRef][Web of Science][Medline]
Nabors L, Lehmkuhl H, Christos N, Andreone TL. Children with diabetes: Perceptions of supports for self-care at school. Journal of School Health (2003) 73:216–221.[Web of Science][Medline]
Sass DA, Smith PL. The effects of parceling unidimensional scales on structural parameter estimates in structural equation modeling. Structural Equation Modeling (2006) 13:566–586.[CrossRef]
Sato AF, Berlin KS, Hains AA, Davies WH, Smothers MK, Clifford L, et al. Teacher support of adherence for adolescents with type 1 diabetes: Preferred teacher support behaviors and youth's perceptions of support. In: The Diabetes Educator. (in press).
Satorra A, Bentler PM. Corrections to test statistics and standard errors in covariance structure analysis. In: Latent variables analysis: Applications for developmental research—Von Eye A, Clogg CC, eds. (1994) Newbury Park, CA: Sage Publications. 399–419.
Siminerio LM, Koerbel G. A diabetes education program for school personnel. Practical Diabetes International (2002) 17:174–177.[CrossRef]
Thomas AM, Peterson L, Goldstein D. Problem solving and diabetes regimen adherence by children and adolescents with IDDM in social pressure situations: A reflection of normal development. Journal of Pediatric Psychology (1997) 22:541–561.
Ullman JB. Structural equation modeling. In: Understanding multivariate statistics—Tabachnick BG, Fidell LS, eds. (2001) 4th. Needham Heights, MA: Allyn & Bacon. 653–771.
Wagner J, Heapy A, James A, Abbott G. Brief report: Glycemic control, quality of life, and school experiences among students with diabetes. Journal of Pediatric Psychology (2006) 31:764–769.
Wysocki T, Greco P, Buckloh LM. Childhood diabetes in psychological context. In: Handbook of pediatric psychology—Roberts MC, ed. (2003) 3rd. New York: Guilford. 304–320.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||