http://dx.doi.org/10.20511/pyr2018.v6n2.212

RESEARCH ARTICLE

Validation of motivational factors for sports activities in university students through SEMPLS

Validación de factores motivacionales para actividades deportivas en alumnos universitarios mediante SEMPLS

Summary

The objective of this research work is to evaluate the scale of motivational factors (appearance, competence, enjoyment, fitness and social) for sport activities using SEMPLS multivariate methods in order to verify the importance of the reliability and validity of the measurement model, based on a construct. The sample usedconsisted of 401 university students from administrative sciences programs and related programs of a Peruvian university. The results found report an average variance extracted of 69.5% for each of the variables and values below the point 0.90 (moderate) for the second order discriminant validity for the HTMT criterion (Heterotrait -Monotrait).

Keywords: Motivational factors, physical activities, structural equations

Resumen

La presente investigación tiene como objetivo evaluar la escala de factores motivacionales (apariencia, competencia, disfrute, fitness y social) hacia las actividades deportivas utilizando las técnicas de métodos multivariados SEMPLS con la finalidad de comprobar la importancia de la fiabilidad y validez del modelo de medida, basado en un constructo. La muestra utilizada de 401 alumnos universitarios de las carreras de ciencias administrativas y carreras afines de una universidad peruana. Los resultados encontrados reportan una varianza extraída media de 69.5% para cada una de las variables; la validez discriminante de segundo orden para el criterio HTMT (Heterotrait–Monotrait) con valores debajo del punto 0.90 (moderado).

Palabras clave: Factores motivacionales, Actividades físicas, Ecuaciones estructurales

Introduction

The motivational theory is studied from different approaches or perspectives. One of them is the self-determination theory, which manifests itself in the autonomy of a person to choose and direct his actions (Deci & Ryan, 1985, 1991; Moreno & Martínez, 2006; Ryan & Deci, 2000). Motivation is determinant of human behavior (Iso-Ahola & St.Clair, 2000). People can be influenced to engage in a certain activity either voluntarily or prompted by social conditioning, to maintain appearances, among other factors, other than for truly essential reasons (Hellín, Moreno, & Rodríguez, 2006).

The studies about motivational theory recognize its multidimensionality. Intrinsic motivation is concerned with activities related to adherence that provides enjoyment and competence. Extrinsic motivation seeks rewards, external aspects to behavior and finally, amotivation (Deci & Ryan, 1985; Weinberg, Tenenbaum, McKenzie, Jackson, Ashel, Grove & Fogarty, 2000). Motivation plays an important role in psychological health, academic performance, decision to stay or drop out of university studies, among others (Del Valle, Matos, Díaz, Pérez & Vergara, 2018).

The decision to start sport activities has been object of many studies about motivation for this activity. According to Tabernero (1998) and Pavón et al. (2008), the most important reasons have been to be physically fit, improve their level, improve their skills, have fun, do exercises and be in shape. Enjoyment is also a factor that persistently increases (Wankel, 1993). Intrinsic motivation is the reason young people consider for sport (Villamarin, Mauri & Sanz, 1998). The studies differentiate the motivation factors for the participation in sport activities according to gender, as that presented by Koivula (1999) in which men lean towards competition; Martínez (2003), Hellín et al. (2004) point out the concern about the body image and the aesthetics mainly in women. Another reason to participate in sport activities is health and lifestyle (Savage, 1998).

It is necessary to infer the motivations of people through behavior or by using several instruments such as personal reports and/or questionnaires. Many studies have validated the scales referred to motivational factors to conduct sport activities. One of them was performed by Frederick and Ryan (1993). They employed a sample of 376 employees from a hospitable university to whom a 23 item questionnaire was applied in order to measure motivation for engaging in physical activities considering enjoyment, competence factors and the factor related to body.The technique used was the exploratory factor analysis with promax rotation and correlation of independent factors.

Ryan et al. (1997) applied a sample of 155 students and workers from a university sport center, in which a 30 items scale was used. These items were grouped in five factors: enjoyment (7 items), appearance (6 items), social (5 items), fitness/health (5 items) and competence (7 items). Enjoyment and competence have been used to reflect an intrinsic motivational orientation, and the other three (social, fitness/health and competence) reflect the external levels of motivation to the person. The polychotomous items are measured in a 7-point Likert scale. A factor analysis with varimax rotation was applied. The Cornbrash’s alpha for each one of them is .92, .88, .83, .78 and .91 respectively. The five factors account for 66% of the total variance. Other authors like Frederick and Schuster-Schmidt (2003); Xu ad Biddle (2000) have used and validated this scale in different contexts and they obtained a good reliability and validity values in all of them.

Moreno et al. (2005) validated for the Spanish context the scale of Ryan et.al. (1997). A sample of 464 people who practice non-competitive physical activities of a Spanish city was used. The sample responds to the statement "Performed physical activity…" through the items making up the scale. The polychotomous items are measured in a 7-point Likert scale where 1 corresponds to "nothing true for me" and 7 "fully true for me." An exploratory factor analysis of main components with direct oblimin rotation was carried out using self-values higher than 1 and a total variance explained of 69.36%.

Then, an internal consistency analysis was performed to determine the reliability coefficient of Cronbach’s alpha. Enjoyment (.84), appearance (.87), social (.81), fitness (.80) and competence (.85). A correlation analysis between all the variables, being positive and significant among all factors. Enjoyment variable had higher correlation with competence and fitness.

The aforementioned studies show evidences of the application of Likert scales where the statistical techniques of factor analysis and correlation were found to be acceptable to measure their reliability and validity. This study is an extension of previous studies that try to evaluate the scale of motivational factors for sport activities of Ryan et.al. (1997), using the SEMPLS multivariate methods in order to verify the measurement model that has the analysis of reliability and validity, considering mainly the construct of the non-observable variables, latent variables.

Method

A descriptive-correlational design was used. The population is composed of university students from the administrative sciences program and related programs of a private university in Lima. Table 1 shows the characteristics of the students in relation to the sport activities they practice. The sample consisted of 401 university students (177 men and 224 women) with ages ranging from 16 to 48: 74.0% are female students and 73.2% are male students, who have attended at the university for less than 3 years. In addition, 49.5% of women and 48.3% of men are in the first two years of studies and more than 50% study in the management and international business schools. Moreover, 60.5% of women and 90.4% of men perform a type of sport and more than 84.3% of women and 93.2% of men perform it voluntarily.

Instrument

The instrument is based on the questionnaire of Ryan et al. (1997), translated by Moreno et al. (2005) that consider 30 items of the factors appearance (6 items), enjoyment (7 items), social (5 items), fitness (5 items) and competence (7 items). (See Table 2); with the 6-point Likert scale having small modifications in the interpretation of the polytomic Likert scale ofMoreno et al. (2005), in which 1 corresponds to "fully disagree" and 6 "fully agree."

Validity and Reliability Analysis

Validity and reliability were carried out through structural equations with partial least squares (SEM-PLS), which is a second generation technique in the area of multivariate methods. This technique was used since the scale used in this study is multidimensional, so the statistical package SmarthPLS (Ringle, Wende & Becker, 2016), a software developed to test structural models, was used.

Structural equations to model a latent construct consider two models: measurement and structural. Reliability and validity of the scales are determined in the measurement model. Reliability was analyzed through two indicators: Cronbach’s alpha and composite reliability, generally the interpretation of the values is the same. The values .7, .8, .9 are considered acceptable, good and excellent, respectively. These two indicators indicate if the scales shows internal consistency. (Hair, Hult, Ringle & Sarstedt, 2017)

As for validity, the convergent and discriminant validity was examined. For the former, the factorial loads must be higher than .708 (authors suggest not to be so rigid in the preliminary stages of the analysis) and the average variance extracted was calculates, in which an acceptable threshold is equal to or higher than 0.5, which means that the latent construct obtained a variance explained of 50%. According to Fornell Larcker criterion, the square root of the AVE of the dimensions should be higher that the correlations with other dimensions in the model, thereby confirming the independence of the latent variables. (Hair et al. 2017).

In addition, the criterion of Heterotrait Monotrait Ratio (HTMT) was used (Henseler et al, 2015) to validate the constructs of each one of the factors, since these scales are evaluated as models of hierarchical components, that is, by levels. According to this indicator, the coefficients must be under the conservative point (.90). Moreover, to calculate the reliability intervals, the Bootstrapping resampling technique (sample= 5000 times, using the option without changing the sign) was used, and they should not include value 1. At the same time, a discriminant validity between the second order construct and dimensions making it up may not be established. However, this result is to be expected, since the measurement model of the construct repeats the indicators of the first order dimensions. (Hair, Hult, Sarstedt, Ringle, & Gudergan, 2017).

Furthermore, the main focus is the second order construct and its measurement model is built from the relationship of the second order construct with its dimensions. In fact, to calculate the average variance extracted and the composite reliability, path coefficients are considered.

Results

Table 3 shows descriptive statistics of motivation for sport activities. It is observed that the items show averages that vary for enjoyment (M=5.11) and appearance (M= 4.48). We can observe that enjoyment and competence show a higher average (M= 5.07), which corresponds to intrinsic motivation. Data show asymmetry and kurtosis within the range -1 to +1, which indicates that they meet the normality assumption.

Figure 1 shows the analysis of constructs of all the factorial loads of the first order dimensions, which represent the motivation factors for sport. All of them show significance above .708, except for item P25 "because I will feel less physically attractive if I do not practice exercises." (.685)

Composite reliability values vary between .915 and .946 (see Table 4). There is internal consistency in such scale. The average variance extracted (AVE) of each dimension of motivational factors for participating in sport activities and for the same construct is satisfactory and varies between .685 and .760.

For the evaluation of the discriminant validity between latent variables, the Fornell-Larker’s criterion as well as the Heterotrait Monotrait (HTMT) criterion are used. The results are evaluated by levels between first order constructs, since the HTMT values must be below the conservative threshold value of .90 (see Table 5). In addition, the reliability intervals do not include the unit, with which this criterion is met.

For the second order, the discriminant validity cannot be established, since the results include first order indicators. Thus, the reliability intervals could not include the unit. The results shown are valid according to these criteria.

Discussion

The dimensions used in this study corroborate the theoretical construct of the motivational factors for sport activities, in which the multidimensionality of motivation is shown through intrinsic and extrinsic reasons (Deci & Ryan, 1985; Weinberg, Tenenbaum, McKenzie, Jackson, Ashel, Grove & Fogarty, 2000). They are determined voluntarily or not by people. (Hellín, Moreno, & Rodríguez, 2006).

It is necessary to mention that the sample used in this study (n=401) has specific characteristics, and consists of university students (homogenous sample). Consequently, the results are more consistent, compared to samples of studies of Ryan et al. (1997) and Moreno et al. (2005), where the former is a sample of students and workers, the latter is a sample of people who practice sport activities in a specific city, they can be understood as more heterogeneous. However, the results found are satisfactory, with a high internal consistency of the scales of motivational factors.

In research work, the motivational factors that best explain motivation are those intrinsic motivational factors (enjoyment and competence), as corroborated in the studies of Ryan et al. (1997) and Moreno et al. (2005), and according to Villamarin, Mauri & Sanz (1998), intrinsic motivation is what young people consider best.

The latent construct modeling, considering the 5 dimensions, reported an average variance extracted for each one of them higher than 50% and for the second order, an average variance extracted (AVE) of 69.5%, this proves that it is satisfactory, consistent with the findings of Moreno et al. (2005) for which an exploratory factor analysis and a total variance explained of 69.36% were used. This SEMPLS technique gives more accuracy in results.

Reliability of each one of the motivational factors of the study shows values ranging from .915 to .946 close to 1, satisfactory, exact and consistent results that provide a stable measurement, they are consistent with what was found by Ryan et al. (1997) and Moreno et al. (2005). The latter showed a Cronbach’s alpha that ranges from .80 to .87. The technique used in this study takes into account the interrelationships of the constructs extracted.

The instrument that shows this research work contains 30 items proposed by Ryan et al. (1997), which using SEM-PLS, show factorial loads higher than 0.685, the scale structure being appropriate. In Moreno et al. (2005) based on the translation of Ryan, eliminated 2 items because they did not have high correlation and did not reach saturation of 0.40, resulting in a 28-item questionnaire according to the factor analysis of the scale, showing, therefore, restrictions in its analysis with respect to this study.

This SEM-PLS approach applied to complex models explains better the relationships between several factors, giving more confidence in its results than the first generation or traditional techniques.

References

Deci, E., & Ryan, R. (1985).The general causality orientations scale: Self- determination in personality. Journal of Research in Personality, 19, 109-134. Doi: https://doi.org/10.1016/0092-6566(85)90023-6

Deci, E., & Ryan, R. (1991). A motivational approach to self: Integration in personality. En R. Dienstbier (Ed.), Nebraska symposium on motivation: Vol. 38. Perspectives on motivation (pp. 237-288). Lincoln, NE: University of Nebraska Press.

Del Valle, M., Matos, L., Diaz, A., Peŕ ez, M.V., & Vergara J. (2018). Propiedades psicométricas escala satisfacción y frustración necesidades psicológicas (ESFNPB) en universitarios chilenos. Propósitos y Representaciones, 6(1), 301- 350. Doi: http://dx.doi.org/10.20511/pyr2018.v6n1.202

Frederick, C., & Ryan, R. (1993). Differences in motivation for sport and exercise and their relations with participation and mental health. Journal of Sport Behavior, 16, 124-146.

Frederick-Recascino, C., & Schuster-Smith, H. (2003). Competition and intrinsic motivation in physical activity: A comparison of two groups. Journal of Sport Behavior, 26, 240-254.

Hair, J., Hult, G., Ringle,C., & Sarstedt, M. (2017). A primer on Partial Least Squaresw Structural Equation Modeling (2nd Ed). Thousand Oaks: Sage.

Hair, J., Hult, G., Ringle,C., Sarstedt, M.& Gudergan, S. (2017). Advanced issues in Partial least Squares Structural Equation Modeling. Los Angeles: Sage.

Hellín, P., Moreno, J., & Rodríguez, P. (2004). Motivos de práctica físico- deportiva en la Región de Murcia. Cuadernos de Psicología del Deporte, 4(1-2), 101-116.

Hellín, P., Moreno, J. A., & Rodríguez, P. L. (2006). Relación de la competencia motriz percibida con la práctica físico-deportiva. Revista de Psicología del Deporte, 15(2), 219-231.

Henseler, J., Ringle, C.M. & Sarstedt, M. (2015). A new criterion for assessing discriminant validity in variance-based structural equation modeling. Journal of the Academic of Marketing Science. 43(1), 115-135. https://doi.org/10.1007/s11747-014-0403-8

Iso-Ahola, S. E., & St.Clair, B. (2000). To- ward a theory of exercise motivation. Quest, 52, 131- 147. Doi: https://doi.org/10.1080/00336297.2000.10491706

Koivula, N. (1999). Sport participation: Differences in motivation and actual participation due to gender typing. Journal of Sport Behavior, 22, 360-380

Martínez, E. (2003). La salud como motivación para la práctica de actividad física en personas adultas. Iatreia, 16(1), 32-43.

Moreno, J., Cervello, E., & Martinez, A. (2005). Validación de la Escala de medida de los motivos para la actividad física –revisada en españoles: Diferencias por motivos de participación. Anales de psicología, 23(1), 167-176.

Moreno, J., & Martínez, A. (2006). Importancia de la Teoría de la Autodeterminación en la práctica físico-deportiva: Fundamentos e implicaciones prácticas. Cuadernos de Psicología del Deporte 6(2), 39-54.

Pavón, A., & Moreno, J. (2008). Actitud de los universitarios ante la práctica físico-deportiva: Diferencias por géneros. Revista Psicología del Deporte 17, 7-23.

Ringle, C., Wende, S. & Becker, J. (2016). SmartPLS 3, Boenningstedt: SmartPLS GmbH. Recuperado de: www.smartpls.com

Ryan, R., Frederick, C., Lepes, D., Rubio, N., & Sheldon, K. (1997). Intrinsic motivation and exercise adherence. International Journal of Sport Psychology, 28, 335-354

Ryan, R., & Deci, E. (2000). Self-determination theory and the facilitation of intrinsic motivation social development, and well--‐being. American Psychologist, 55, 68‐78. Doi: https://doi.org/10.1037/0003-066X.55.1.68

Savage, M. P. (1998). University students’ motivation for participation in a basic instruction program. College Student Journal, 32 (1), 58-65.

Tabernero, B. (1998). Motivos para practicar tenis en la iniciación deportiva. En A. García, F. Ruiz y A. J. Casimiro (Eds.), Actas del II Congreso Internacional sobre la Enseñanza de la Educación Física y el Deporte Escolar (pp. 402-406). Málaga: Instituto Andaluz del Deporte.

Villamarín, F., Maurí, C. & Sanz, A. (1998). Competencia percibida y motivación durante la iniciación en la práctica del tenis. Revista de Psicología del Deporte 13, 41-56

Wankel, L. (1993). The importance of enjoyment to adherence and psychological benefits from physical activity. International Journal of Sport Psychology, 24, 151-169.

Weinberg, R., Tenenbaum, G., McKenzie, A., Jackson, S., Anshel, M., Grove, R., & Fogarty, G. (2000). Motivation for youth participation in sport and physical activity: relationships to culture, self-reported activity levels, and gender. International Journal of Sport Psychology 31, 321-346.

Xu, X., & Biddle, S. (2000). Difference in motivation for adherence, gender and age. Paper presented in Pre-Olympic Congress Sport Medicine and Physical Education International Congress on Sport Science, Brisbane, Australia.