Comparative concurrent validity of the MRF-L and ARS competitive state Anxiety Rating Scales for volleyball and basketball.

Cox, Richard H. ; Russell, William D. ; Robb, Marshall 等

Measurement of competitive state anxiety during and before competition has suffered from a number of measurement-related limitations. Martens (1982) and Landers (1980) disagreed on the best way to measure competitive state anxiety. Martens argued that a pencil and paper assessment of state anxiety was as accurate and reliable as a physiological assessment. Landers, however, argued that both physiological and a psychological assessment is necessary. These two divergent opinions are complicated by the fact that low correlations exist among physiological measures of state anxiety and between psychological and physiological measures (Karteroliotis & Gill, 1987; Landers, 1980). These early observations of Martens and Landers remain as viable today as they were in the early 1980's. Further compounding this complex problem is the observation that measurement of state anxiety prior to competition may be different than state anxiety measured during competition. It is very difficult, however, to get accurate estimates of competitive state anxiety during competition. An exception to this general observation may include individual sports such as golf, diving, and bowling where delays of various lengths occur between strokes, dives, and rolls. It is more difficult to imagine assessing state anxiety during competition involving a team sport such as volleyball or basketball, but if the measurement instrument was unobtrusive, it would be possible.

Spielberger's State Anxiety Inventory (SAI) has been utilized by many researchers to obtain an estimate of precompetitive state anxiety in athletes (Spielberger, 1983; Spielberger, Gorsuch, & Lushene, 1970). In an effort to shorten the time necessary to administer the SAI, Martens (1977) developed the Competitive State Anxiety Inventory (CSAI), which is a shortened sport specific version of the SAI. Use of the CSAI was short lived, however, due to the development of multidimensional tests of state anxiety such as the Competitive State Anxiety Inventory-2 (CSAI-2).

The CSAI-2 provided a measurement of both cognitive and somatic competitive state anxiety (as well as self-confidence). This distinction between cognitive and somatic state anxiety has been pivotal in helping sport psychologists better understand the relationship between anxiety and performance (Cox, 1998). Specifically, this distinction is based on a multidimensional theory of competitive state anxiety from which the CSAI-2 was developed. Multidimensional theory posits that a distinction needs to be made between cognitive state anxiety and somatic state anxiety, and that their relative independence warrants separate measurement (Martens, Burton, Vealey, Bump, & Smith, 1990). While psychometrically sound, the CSAI-2 is 27 items in length and does not lend itself to measurement of anxiety immediately prior to an at bat in softball or a stroke in golf. The typical administration period for the CSAI-2 is 15-30 minutes before competition and the estimated test completion time is approximately 5 minutes. While this time frame is sufficient for precompetitive measures, it is rare that the athlete has this time available during competition for assessment of state anxiety.

Hanin (1986) proposed a retrospective method of determining state anxiety. Using retrospective techniques athletes are asked to recall their level of state anxiety several days after a specific event. If reliable, this technique would allow researchers to unobtrusively obtain precompetitive state anxiety. Harger and Raglin (1994) reported that collegiate track and field athletes can accurately recall precompetitive state anxiety two days after the initial event. It is not known if the retrospective technique could be utilized to recall state anxiety during competition. To clarify the relationship between state anxiety and performance, McAuley (1985) made an early recommendation that future researchers attempt to assess state anxiety during competition. In an effort to address this recommendation, Murphy, Greenspan, Jowdy, and Tammen (1989) developed the Mental Readiness Form (MRF). As an alternative to the CSAI-2, the MRF was designed to assess precompetitive state anxiety through a 3-item instrument employing a visual analog scale. The three parts of the MRF correspond to cognitive anxiety, somatic anxiety, and self-confidence subscales of the CSAI-2. Krane (1994) replaced the visual analog scale in the original MRF with an 11-point Likert scale and called it the MRF-Likert (MRF-L). Krane also developed the MRF-3 which used the bipolar opposites of worried-not worried, tense-not tense, and confident-not confident to represent cognitive anxiety, somatic anxiety, and self-confidence. Krane concluded that all three versions of the MRF exhibit concurrent validity when compared with the CSAI-2, but that the MRF-L and MRF-3 versions were preferred to the original MRF because of clarity of response gained when using a Likert scale. Her data also revealed higher correlations with CSAI-2 subscales when the MRF-L was used as opposed to either the MRF or the MRF-3. She reported a simple correlation of .76 between the MRF-L and corresponding CSAI-2 cognitive subscale and .69 between the MRFL and the corresponding CSAI-2 somatic subscale.

While the Mental Readiness Scale has emerged as a viable alternative to the CSAI-2 when an unobtrusive measurement tool is desired, Krane (1994) cautions that it should not be considered a viable replacement for the more psychometrically sound CSAI-2. It should also be pointed out that the MRF was not developed in any psychometric sense from the CSAI-2 and cannot be technically considered a short form of the CSAI-2. The anchors, however, as utilized in both the MRF-L and MRF-3 appear to be theoretically selected to represent the measurement of cognitive anxiety, somatic anxiety, and self-confidence.

Somewhat concurrent with the development of the MRF-L and the MRF-3, Cox, Russell, and Robb (1998) reported the development of the Anxiety Rating Scale (ARS). The ARS was developed directly from the CSAI-2 and utilizes two brief statements from the parent test to measure cognitive and somatic anxiety. The ARS was developed using stepwise multiple regression techniques and follow-up concurrent validity checks with the parent CSAI-2 test. Both the MRF-L and ARS independently exhibit concurrent validity coefficients with the CSAI-2 in the magnitude of .60 to .84. It is not known, however, which of the two rating scales would be most predictive of CSAI-2 subscales in the same competitive environment with the same subjects. The purpose of the present investigation, therefore, was to compare the concurrent validity of the MRF-L and ARS in the same competitive environment with the same participants. Because the ARS was originally developed and tested utilizing a sample of 771 volleyball and basketball intramural athletes, it was hypothesized that the current independent sample of volleyball and basketball athletes would yield validity coefficients favoring the ARS over the MRF-L.

Method

Participants

Participants for this research were 418 male and female intramural athletes competing in volleyball or basketball at a large Midwest University. Two-hundred and twenty-four of the subjects competed in volleyball while 194 competed in basketball. The average age of the 211 male athletes was 20.3 yr (SD = 1.96), while the average age of the 207 female athletes was 19.2 yr (SD = 1.11). Use of human subjects for the purpose of research was approved by the appropriate University human subject's committee and anonymity assured.

Instruments

The instruments used for collecting anxiety data were the Competitive State Anxiety Inventory-2 (CSAI-2), the Anxiety Rating Scale (ARS), and the Mental Readiness Form (MRFL). The CSAI-2 is composed of 27-items that measure cognitive state anxiety, somatic state anxiety, and self-confidence. Details associated with the development and testing of the CSAI-2 are reported by Martens, Vealey, and Burton (1990).

The MRF-L (Krane, 1994) is composed of three brief statements associated with an 11 point Likert scale. The three brief statements, associated with cognitive anxiety, somatic anxiety, and self-confidence respectively, are "my thoughts are", my body feels", and "I am feeling." The anchors for the "my thoughts are" cognitive anxiety statement were calm and worried. The anchors for the "my body feels" somatic anxiety statement were relaxed and tense. Finally, the anchors for the "I am feeling" self-confidence statement were confident and scared. A high score for self-confidence in the CSAI-2 indicates high self-confidence, whereas a high score for self-confidence on the MRF-L indicates low self-confidence. Thus, the two scales will be negatively correlated.

The ARS (Cox et al., 1998) is composed of three brief statements associated with a seven point Likert scale. The three brief statements in the ARS are as follows:

Cognitive Anxiety - I feel concerned about performing poorly and that others will be disappointed with my performance.

Somatic Anxiety - I feel nervous, my body feels tight and/or my stomach tense.

Self-Confidence(1) - I feel secure, mentally relaxed, and confident of coming through under pressure.

The Likert scale associated with each of these statements runs from 1 (not at all) to 7 (intensely so). Each statement represents an aggregate of three items from the CSAI-2. The three items that make up each statement were selected on the basis of maximum-[R.sup.2] (SAS) stepwise multiple regression procedures. Thus, the three selected items (out of nine) were those items that together contributed the most to predicting the CSAI-2 subscale for cognitive anxiety, somatic anxiety, or self-confidence. In each case, the three selected items represented the best three variable model in terms of maximizing the squared multiple correlation. Intercorrelations among the three selected somatic anxiety items ranged from .441 to .545, while for cognitive anxiety and self-confidence the range was .215 to .517 and .409 to .505 respectively. While issues related to collinearity are of concern when analyzing effects, they pose little problem when the researcher's sole purpose is in the determination and interpretation of the squared multiple correlation (Pedhazur, 1997). While the concurrent validity of the ARS is being addressed in the current investigation, the test-retest reliability of the ARS was reported to be .97 for cognitive state anxiety, .90 for somatic state anxiety, and .95 for self-confidence (Cox, Davis, & Robb, 1998).

Procedures

During round-robin and play-off volleyball and basketball competition, starting members of selected teams completed the MRF-L, ARS, and CSAI-2 approximately 15 minutes before match/game. During round-robin competition, records were kept so that selected teams were tested only once. During the play-off games, it was not possible to avoid a repeated measurement situation, in which an athlete was measured more that once. Measurements were considered to be independent of each other, however, because of the varying conditions that prevail prior to each game (Turner & Raglin, 1996). The order in which the MRF-L and ARS were administered to athlete's was counterbalanced to eliminate an order bias. For example, of the I 13 male volleyball players tested, approximately half of them completed the MRF-L before the ARS and the other half completed the ARS before the MRF-L. The CSAI-2 was always administered last. Participants were briefed regarding the nature of inventories, asked to read questions carefully and respond in like manner. Observations were obtained from 117 males prior to round-robin games, 110 females prior to round-robin games, 94 males prior to play-off games, and 97 females prior to play-off games.

Statistical Analyses

Correlation techniques were utilized to determine the relative concurrent validity of the ARS and MRF-L with the CSAI-2. A t-test for comparing two dependent correlations was utilized to compare r's associated with the ARS and the MRF-L (Bruning & Kintz, 1987). In order to ascertain the influence of gender, situation (round-robin, play-off), and sport upon the ability of the ARS or MRF-L to predict subscales associated with the CSAI-2, they were included in multiple regression analyses.

Results

Correlational relationships between MRF-L and CSAI-2 subscales, and between ARS and CSAI-2 subscales are illustrated in Table 1. As can be observed in Table 1, correlation results are compared for all 418 subjects combined, but also for sub-groups as a function of gender, situation (round-robin vs. play-off games), and sport. Sub-group breakdowns provide insight into comparative relationships. In the case of cognitive state anxiety, the ARS consistently exhibits higher correlations with the CSAI-2 subscale for cognitive anxiety than does the MRF-L. The single exception to this observation is in the play-off game situation. In the case of somatic state anxiety, the ARS again consistently exhibits higher correlations with the CSAI-2 subscale for somatic anxiety than does the MRF-L. The single exception to this observation was in the case of the female participants. In the case of self-confidence, however, the MRF-L exhibits higher correlations with the CSAI-2 subscale for self-confidence than does the ARS. The single exception to this observation was in the case of male intramural athletes.

Also noted in Table 1 are the results of dependent t-tests between ARS and MRF-L correlations with the CSAI-2. The results of these comparisons revealed five significant differences at the .05 level and one at the .01 level. All but one of these differences favored the ARS over the MRF-L. The differences for cognitive and somatic anxiety favored the ARS, while the single difference for self-confidence favored the MRF-L.

In order to ascertain the influence of gender, situation (round-robin, play-off), and sport upon the ability of the MRF-L or ARS to predict competitive state anxiety as measured by the CSAI-2, they were included in a multiple regression analysis. Utilizing either the MRF-L or the ARS as the single continuous property predictor variable, and gender, situation, and sport as categorical predictor variables, six separate regression models were tested as illustrated in [TABULAR DATA FOR TABLE 1 OMITTED] Table 2. These regression analyses provide insight as to how gender, situation, and sport interact with the MRF-L or ARS in predicting subscales of the CSAI-2. As can be observed in Table 2, the MRF-L or ARS scores were significant predictors of subscales of the CSAI-2 in all six models. Gender was revealed to be a significant parameter for predicting somatic state anxiety for both the ARS and the MRF-L. As gender switches from male to female, either a 1.57 (MRF-L) or 1.12 (ARS) point reduction in somatic anxiety is predicted (as measured by the CSAI-2).

Discussion

As previously stated, the purpose of this investigation was to compare the concurrent validity of the MRF-L and ARS in the same competitive environment with the same participants. Because the ARS was originally developed and tested utilizing a sample of volleyball and basketball intramural athletes, it was hypothesized that the current independent sample of volleyball and basketball athletes would yield validity coefficients favoring the ARS over the MRF-L.

Results revealed that for all participants combined, the ARS is a stronger predictor of cognitive and somatic state anxiety as measured by the CSAI-2, while for confidence, the MRF-L is a stronger predictor. based upon the results of this study it may be concluded that the ARS provides a more accurate assessment of state anxiety in basketball and volleyball players than does the MRF-L. Use of the ARS will make it possible to ascertain an athlete's precompetitive state anxiety at a point in time closer to the actual event, and to measure state anxiety during actual competition. Krane, Joyce, and Rafeld(1994) and Durr (1996) provide illustrations of how this can be accomplished in research involving softball batters and 3-meter board divers.

Development and testing of short inventories capable of accurately measuring anxiety and self-confidence at a point in time contiguous with actual performance seems critical in the ultimate understanding of the effects of affect on performance. Much work still remains to be done, however. The current investigation suggests that the ARS is more accurate than the MRF-L relative to the measurement of somatic and cognitive state anxiety in selected team sport athletes (concurrent validity). Results of research reported by Cox, Reed, and Robb (1996) suggest that the same is true for team sport athletes.

The construct validity of the ARS must also be addressed. Athletes, for example, who have been exposed to psychological skills training should be better able to control their state anxiety prior to an important competition. The ARS, or for that matter the MRF-L, should be able to distinguish between a group of athletes who have experienced psychological skills training and a group who has not.

[TABULAR DATA FOR TABLE 2 OMITTED]

Note

1 The Cox et al. (1998) article chronicling the development of the ARS did not include measurement of self-confidence, although data were collected on this variable. Results of a stepwise regression procedure identified three CSAI-2 self-confidence items that together accounted for 86% of the variability of the CSAI-2 self-confidence subscale. These three were items 12 (I feel secure), 21 (I feel mentally relaxed), and 27 (I'm confident of coming through under pressure). As with cognitive and somatic anxiety, these three items were condensed into a single aggregated statement and set to a 7-point Likert scale in the current research.

References

Bruning, J.L., & Kintz, B.L. (1987). Computational handbook of statistics (3rd ed.). Glenview, IL: Scott Foresman.

Cox, R.H. (1998). Sport psychology: Concepts and applications (4th ed.). Dubuque, IA: WCB/McGraw-Hill.

Cox, R.H., Russell, W.D., & Robb, M. (1998). Development of a CSAI-2 Short form for assessing competitive state anxiety during and immediately prior to competition. Journal of Sport Behavior, 21, 30-40.

Cox, R.H., Reed, C., & Robb, M. (1996). Comparative concurrent validity of the MRF-L and ARS competitive state anxiety rating scales for individual sport athletes. Journal of Applied Sport Psychology, 8, Supplement, S23.

Cox, R.H., Davis, J.E., & Robb, M. (1998). Test-retest reliability of the anxiety rating scale. Research Quarterly for Exercise and Sport, 69, Supplement, Abstract 111.

Durr, K.R. (1996). Relationship between state anxiety and performance in high school divers. Unpublished master's thesis, University of Missouri, Columbia.

Hanin, Y.L. (1986). State - trait anxiety research on sports in the U.S.S.R. In C.D. Spielberger, & R. Dias-Guerrero (Editors), Cross - cultural anxiety (pp. 45 - 64). Washington D.C.: Hemisphere Publications.

Harger, G.J., & Raglin, J.S. (1994). Correspondence between actual and recalled precompetition anxiety in collegiate track and field athletes. Journal of Sport & Exercise Psychology, 16, 206-211.

Karteroliotis, C., & Gill, D.L. (1987). Temporal changes in psychological and physiological components of state anxiety. Journal of Sport Psychology, 9, 261-274.

Krane, V. (1994). The mental readiness form as a measure of competitive state anxiety. The Sport Psychologist, 8, 189-202.

Krane, V., Joyce, D., & Rafeld, J. (1994). Competitive anxiety, situation criticality, and softball performance. The Sport Psychologist, 8, 58-72.

Landers, D.M. (1980). The arousal - performance relationship revisited. Research Quarterly for Exercise and Sport, 51, 77-90.

Martens, R. (1977). Sport competition anxiety test. Champaign, IL: Human Kinetics Publishers.

Martens, R. (1982). Sport competition anxiety test(2nd ed.). Champaign, IL: Human Kinetics Publishers.

Martens, R., Burton, D., Vealey, R.S., Bump, L.A., & Smith, D. (1990). Development and validation of the competitive state anxiety inventory-2. In R. Martens, R.S. Vealey, & D. Burton (Eds.), Competitive anxiety in sport (pp. 117-190). Champaign, IL: Human Kinetics Books.

Martens, R., Vealey, R.S., & Burton, D. (1990). Competitive anxiety in sport. Champaign, IL: Human Kinetics Publishers.

McAuley, E. (1985). State - anxiety: Antecedent or result of sport performance. Journal of Sport Behavior, 8, 71 - 77.

Murphy, S.M., Greenspan, M., Jowdy, D., & Tammen, V. (1989). Development of a brief rating instrument of competitive anxiety: Comparisons with the Competitive State Anxiety Inventory-2. Proceedings of the Association for the Advancement of Applied Sport Psychology (p.82). Seattle, WA.

Pedhazur, E.J. (1997). Multiple regression in behavioral research (3rd ed.). New York: Harcourt Brace.

Spielberger, C.D. (1983). Manual for the state-trait anxiety inventory (Form-Y). Palo Alto, CA: Consulting Psychologists Press.

Spielberger, C.D., Gorsuch, R.L., & Lushene, R.F. (1970). Manual for the state-trait anxiety inventory. Palo Alto, CA: Consulting Psychological Press.

Turner, P.E., & Raglin, J.S. (1996). Variability in precompetition anxiety and performance in college track and field athletes. Medicine and Science in Sports and Exercise, 28, 378-385.