Reflected Appraisals, Academic Self-Perceptions, and Math/Science Performance During Early Adolescence (Bouchey & Harter, 2005)

Bouchey, H. A., & Harter, S. (2005). Reflected Appraisals, Academic Self-Perceptions, and Math/Science Performance During Early Adolescence. Journal of Educational Psychology, 97(4), 673-686. doi:10.1037/0022-0663.97.4.673

The authors investigate whether or not middle school students’ reflected appraisals of competence, support, and importance were linked to their own corresponding self-perceptions. More specifically, do adolescents’ self perceptions mediate the relations between their perceptions of parents’, teachers’, and peers’ views, behavior, and academic perceptions?

There is an extensive body of research on links between self-perceptions and school achievement (see Eccles et al., 1998), i.e., when students deemed it important to do well in math/science, felt capable of doing well, and reported engaging in behaviors linked to doing well, they would in fact demonstrate successful math/science performance. As individual self-perceptions represent the most proximal influences on academic success, the authors hypothesized that:

  • the relations between adolescents’ perceptions of what others thought about the importance of math/science on academic outcomes would occur indirectly through students’ own beliefs about importance and scholastic behavior.
  • the effects of perceived support from others and adolescents’ perceptions of others’ beliefs about their competence in math/science would also occur indirectly through students’ own perceived competence and scholastic behavior in math/science.

Theoretical basis: Cooley’s (1902) and Mead’s (1934) symbolic interactionist theories (the self is created through the internalization of others’ beliefs about oneself (see Harter, 1998, 1999); Eccles’s (1993) expectancy-value theory (both children’s expectations for doing well in a particular domain and the value they place on doing well in that domain affect their subsequent academic choices and performance).

The role of parents and teachers as socializers of achievement beliefs is also central to expectancy-value theory. Children’s perceptions of socializers’ beliefs, expectations, and attitudes predict their own self-concept (Eccles, 1984; Eccles et al., 1985; Meece, Parsons, Kaczala, Goff, & Futterman, 1982). However, the role of adolescents’ perceptions of others’ beliefs and behavior in predicting their own academic self-perceptions and performance  is relatively unexplored.

The authors proposed a model in which adolescents’ reflected appraisals of what parents, teachers, and peers think about them with respect to math/science would predict their own academic self-perceptions which would in turn predict academic performance.

Achievement in Latino Youth
On average, economically disadvantaged, African American, and Latino adolescents typically fare worse academically (Byrnes, 2003; Henderson, 1997; Patterson, Kupersmidt, & Vaden, 1990; Pungello, Kupersmidt, Burchinal, & Patterson, 1996). Graham’s (1994) seminal review revealed a striking paucity of theoretically grounded research on the processes underlying achievement in African American youth (see all McLoyd, 1998).

The authors expected that Latino students would demonstrate lower mean levels of reflected appraisals, self-perceptions, and math/science performance than would European American students. A novel contribution of this study is examining whether Latino students believe that others think they are less capable at math/science, that others think math/science is less important for them, and that others provide less support for them to do well in math/science — these unfavorable reflected appraisals should predict lower self-perceptions, which should in turn predict lower math/science performance.

These findings could thus shed light on why Latino students, as a group, typically perform less well academically than do European American students.

The effects of SES were controlled for in the current study. “As many scholars have argued, the confounding of socioeconomic factors with ethnic background has stymied progress of the knowledge of achievement processes (Graham, 1994; Stevenson, Chen, & Uttal, 1990). Parental education level is closely linked to children’s math/science achievement (see Byrnes, 2003; Peng et al., 1995). We thus controlled for SES in the form of parental education level in the current study.”


  • Modified scales from the Self-Perception Profile for Adolescents (SPP-A; Harter, 1988) and the Social Support Scale for Older Children and Adolescents (Harter & Robinson, 1988) were used.
  • To minimize response bias, both positively and negatively worded items were included in all scales. Students circled their responses to each item.
  • Mean-level ratings (obtained when the student reported on at least 75% of the items) were used for all observed scales.

Reflected Appraisals
Adolescents’ perceptions of significant others’ beliefs regarding how important it was for them to do well in math/science were assessed with a 7-item measure adapted from the How Important These Things Are scale of the SPP-A. Items from the Academic subscale were modified to assess adolescents’ perceptions of others’ beliefs and to inquire specifically about math/science schoolwork. Participants responded separately to each item for mothers, fathers, teachers, and classmates.

Sample items included “My mother (father, teacher, classmates) think(s) that doing well on tests in math/science is very important” and “My mother feels that getting good grades in math/science is not that big of a deal” (reverse coded).

Support for schoolwork.
Students’ perceived support for math/science schoolwork was assessed with a 10-item scale adapted from the Social Support Scale for Older Children and Adolescents (Harter & Robinson, 1988). Five items from both the Approval Support and the Instrumental
Support subscales were modified to reflect support specifically tailored to math/science.

Sample items included “My mother (father, teacher, classmates) praise(s) me for my schoolwork in math/science” (approval support) and “My mother does not teach me about the things I need to know in math/science” (instrumental support, reverse coded). We created average ratings of perceived support from mothers, fathers, teachers, and classmates using all 10 of the support items for each social partner.

Beliefs about the target student’s competence.
Adolescents’ assessments of significant others’ beliefs about their academic competence were measured with a five-item scale adapted from the What I Am Like subscale of the SPP-A (Harter, 1988). Sample items included “My mother (father, teacher, classmates) believe(s) that I am smart for my age in math/science” and “My mother thinks I am pretty slow at finishing my work in math/ science” (reverse coded).

Importance of schoolwork. Adolescents’ own perceptions of the importance of math/science schoolwork were assessed using modifications of the Academic subscale items from the How Important These Things Are scale of the SPP-A (Harter, 1988). Five items tapping the importance of doing well on achievement tests, assignments, and class tests were modified to inquire specifically about math/science subjects. Sample items included “I think that it is important for me to do well in math/science” and “I feel that getting good grades in math/science is not that big of a deal” (reverse coded).

Scholastic behavior. Seven items assessed the extent to which students engaged in a range of behaviors devoted to specific coursework in math/science. Sample items included “I almost always complete my homework on time in math/science” and “I don’t really put as much time and energy into doing my math/science schoolwork as I should” (reverse coded).

Perceived competence. Adolescents’ perceived academic competence was assessed using modifications of the five Academic subscale items from the What I Am Like scale of the SPP-A (Harter, 1988). Sample items for this scale included “I am smart for my age in math/science” and “I am pretty slow at finishing work in math/science” (reverse coded).

Academic Performance
Current performance. School records of students’ current marking period grades assessed current performance. Letter grades for both math and science classes were converted to a 13-point scale (i.e., A+ = 12, F  = 0).

Prior achievement. Because students’ course grades from the previous grade in school were unavailable, we used school records of students’ achievement test scores from the previous spring (one year before this study) as a measure of prior achievement (percentile ranks from math and science subtests of the Iowa Test of Basic Skills, ITBS).

Parental Educational Level
Students’ reports of their mothers’ highest educational level were used as an index of SES in the current study. Students checked the highest level of education completed by their mothers on a 10-point scale ranging from 1  no formal education to 10  graduate degree. Obtained scores from students ranged from 1 to 10.



“Correlational analyses revealed that adolescents’ perceptions of mothers’, fathers’, and teachers’ beliefs and behavior, but not of classmates’ belief and behavior, were most strongly related.”

  • Students’ perceptions of how important adults thought it was to do well were positively related to their own perceived importance of math/science (rs = .33–.37) and to the effort (behavior) they put forth to succeed in these courses (rs = .31–.35), as well as to their perceived competence in math/science, to a lesser extent (rs =  .15–.27).
  • Perceptions of classmates’ importance beliefs were significantly related only to students’ perceived importance and behavior (rs = .23 -.13, respectively), not to their perceived math/science competence.
  • Reflected appraisals of importance for adults, but not for classmates, were also positively related to students’ grades in math/science (rs = .16 –.21).
  • Moderate correlations for support from mothers, fathers, and teachers with students’ perceived competence and scholastic behavior in schoolwork (rs = .22–.44).
  • Perceived support from classmates was positively correlated with students’ scholastic behavior (r = .21) but not with their perceived competence in math/science.
  • Perceived support from others and students’ perceived importance of math/science was positively correlated (rs = .20 –.51).
  • Perceived support was also positively correlated with math/science performance (rs =  .13–.35).
  • Students’ reflected appraisals of competence for mothers, fathers, teachers, and classmates were significantly correlated with their own perceived competence ratings (rs = .44–.62).
  • A similar pattern of results was found for their academic behavior, although the magnitude of these correlations was slightly smaller (rs  .32–.40).
  • Students’ reflected appraisals of others’ competence beliefs were positively related to their self-rated importance of doing well in these subjects (rs  .18–.32).
  • Students’ reflected appraisals of competence were also positively correlated with their math/science grades (rs  .29 –.37).
  • The more important students thought it was to do well in math/science, the more time and effort they spent to succeed in math/science coursework, and the more competent they felt in these classes, the higher were their math/science grades (rs = .26 –.47).

Structural Equation Models
The test of the model as initially specified it yielded acceptable goodness-of-fit statistics. However, inspection of the modification indices and standardized path
coefficients suggested that five new pathways would significantly improve the fit of the model. Each added pathway statistically improved the fit of the model (as indicated by the modification index for that path) but also made both logical and conceptual sense given the underlying theory of the model.

These new pathways indicated that students’ self-perceptions of both (a) the importance of math/science schoolwork and (b) their competence in these courses predicted their scholastic behavior (time and effort invested in doing well), (c) students’ perceived support for math/science predicted their own importance ratings for these courses, (d) students’ perceptions of adults’ beliefs about their competence in math/science predicted their own importance ratings for these courses, and (e) students’ perceptions of support from teachers had a direct effect on their performance in math/science above and beyond the indirect effect of the latent reflected-appraisal support construct. In addition, because it appeared that the effects of students’ valuing of math/science on academic outcomes occurred indirectly through the amount of time and energy they invested in these courses, we deleted the direct path between self-perceived importance and academic performance in the model.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: