Secondary Logo

Journal Logo

Bottom-up Effects Of Perceived Competence In Sport On General Self-esteem: 942 Board #164 9:00 AM – 10:30 AM

Conroy, David E.; Coatsworth, Douglas

Medicine & Science in Sports & Exercise: May 2005 - Volume 37 - Issue 5 - p S181
C-31: Free Communication/Poster – Psychology in Athletics: THURSDAY, JUNE 2, 2005 9:00 AM - 12:00 PM, OOM: Ryman C2

Penn State University, University Park, PA.

(Sponsor: W. Larry Kenney, FACSM)


Perceived competence (PC) is often conceptualized as a lower-order expression of general self-esteem. Both perceived competence in a specific athletic domain and general self-esteem increase as a function of sport participation; however, dynamic relations between these constructs have not been examined. As such, it is not clear to what extent bottom-up effects of increased PC account for increases in general SE.

Back to Top | Article Outline


To determine the degree to which initial PC and the rate at which PC changes account for inter-individual variability in the rate at which SE changes over a 6-week youth sport season.

Back to Top | Article Outline


Young female (n = 66) and male (n = 69) athletes completed measures of perceived swimming competence (weeks 1, 3, 6) and self-esteem (weeks 1, 4, 6) throughout a summer youth swim league. Longitudinal factorial invariance analyses were conducted to establish measurement equivalence across assessments for each construct. Second-order latent growth curve (LGC) models were employed to test relations between second-order intercept and slope factors for perceived competence self-esteem.

Back to Top | Article Outline


Models of perceived swimming competence and general self-esteem responses exhibited strict and strong longitudinal factorial invariance, respectively. A series of nested LGC analyses revealed that model fit decreased significantly when paths between initial PC and SE were constrained to zero (Δχ2[1] = 67.4, p <.01), and when paths between PC and SE slopes were constrained to zero (Δχ2[1] = 6.0, p <.05). Fit did not decrease significantly when the path between initial PC and SE slopes was constrained to zero (Δχ2[1] = 0.1, p >.05). The correlation between intercept factors was .80 (p <.01) and the standardized regression coefficient for PC slopes predicting SE slopes was 0.69 (p <.05); this model accounted for 48% of the variance in SE slopes.

Back to Top | Article Outline


PC and SE appear to be functionally linked such that changes in domain-specific PC have bottom-up effects on changes in general SE. Supported by NICHD Grant R03 HD42535.

©2005The American College of Sports Medicine