Careful what you say to yourself: Self-talk and tennis performance via multi-level modeling

• In Event: Poster Session 03
  In Poster Session: PS 03 Topic - Social, Emotional, Personality

Abstract

Self-talk helps young athletes to regulate their performance. Self-talk in tennis predicts competitive performance in adult athletes (Van Raalte et al., 2000; Zourbanos et al., 2015). Prior work has not examined how self-regulatory speech and performance are inter-related in the young athlete. The current study extends prior work with young athletes, by examining speech and performance as inter-related with a within-person nested model. We hypothesized the use of negative self-talk would relate to losing points, while positive and instructional self-talk would be linked with winning points. We demonstrate that nested modeling can predict subsequent point performance from self-talk occurring on the prior point, controlling for match length and earlier performance in the match.

Twenty-eight youth tennis players participated. The sample was majority female (64%), from various racial backgrounds (21% Latinx, 29% Asian, 18% White, 25% other), and middle to upper-income families. The setting was a local competitive camp directed by area coaches. We observed players for one set of tennis competition. Coders used the Self-Talk and Gestures Rating Scale (Van Raalte et al., 1994). Audible self-talk was coded on each point as positive (“Great shot!”), negative (“You are terrible!”), instructional (“Swing down on the ball”), or unintelligible.

Unintelligible self-talk occurred on average during 19% of the points played during the match. Negative self-talk occurred on average during 15% of the points observed. Positive (5% of points) and instructional self-talk (1% of points) occurred infrequently. Players completed an average of 8 games per set but some went longer (max of 13 games with a tie-breaker). Table 1 displays the frequency of self-talk (positive, negative, instructional, muttering) cross-tabulated with performance (points won/lost). Negative self-talk tended to occur during or just after losing a point (K), and positive speech during or just after winning that point. No bivariate associations of speech on K and performance at K+1 were found.

We then analyzed (Table 2) the data using logistic multi-level modeling. Level 1 represented the associations within the set, and Level 2 represented associations between players. Negative self-talk increased the odds of winning the concurrent point (B = 2.44 p < .001, OR = .09), and a trending decrease in odds of winning point K+1 (p = .08). Positive self-talk increased the odds of winning the concurrent point (B = 3.37 p < .001, OR = 29.26). Players using positive self-talk on the concurrent point saw a 2%

Authors

• Jordan Thibodeaux, University of Louisiana at Lafayette
  Presenting Author

• Adam Winsler, George Mason University
  Non-Presenting Author