Abstract
Introduction The importance of a multidimensional conceptualization of motivation in sport and Physical Education (PE) proposed by the Self-Determination Theory (Deci & Ryan, 2000) has been supported by a large body of empirical evidence. Recently there was an increased interest in using person-centered analytical approach to examine sport motivational profiles, which reflect the simultaneous presence and interactions of multiple motivations within the same individual.
Although a measurement invariance (e.g., scalar invariance) of the extracted profiles constitutes an important methodological prerequisite of valid between-profile comparisons in outcomes, this prerequisite was never empirically tested in previous studies. Method The Czech version of the Sport motivation scale (Komarc et al., 2020) was used to measure different types of motivational regulations (intrinsic, identified, introjected, external, amotivation) in a sample of N=456 undergraduate students.
The optimal number of profiles was determined using the Latent Profile Analysis (LPA), a model-based clustering method. Invariance tests based on gender, competitive level and type of sport (individual vs. team) were then conducted using a multi-group analysis framework.
Results The model fit indices and parameters for different LPA models supported existence of 4 distinct latent motivational profiles in our sample. The largest latent profile (Moderately autonomous) consisted of 51.5% participants, who reported the highest levels of intrinsic motivation compared to other profiles.
Participants in the smallest profile (Highly amotivated, 7.5%) on the other hand, reported extremely high levels of amotivation. Multigroup LPA supported configural invariance (number of profiles is the same) for all 3 grouping variables.
Further analyses have indicated that a partial metric invariance is a plausible hypothesis after removing equality constrains on several latent profile indicators. Discussion Consisted with our expectations and previous research using the LPA modeling framework, we identified four distinct sport motivation profiles.
Evidence of partial metric invariance for the extracted profiles provides a sound basis for assessing between-profile differences in motivational outcomes, which are of a great interest for coaches and sport psychologists. Future studies are however warranted in order to replicate our findings and/or expand available evidence about invariance in other contexts (e.g., PE) and for other important grouping variables such as age, experience with sport participation, or culture.