test interval of eighteen days (Smith et al., 1990). Evidence of convergent validity for this
scale was reported by Smith et al. (1990) who calculated significant correlations (ranging
between 0.47 and 0.81) between its sub-scales and the Sport Competition Anxiety Test
(SCAT; Martens, 1977). Discriminant validity for the SAS is supported by evidence of
low correlations between it and general mental health measures (see Smith et al., 1998).
Factor analyses have also confirmed that the SAS assesses three separate dimensions:
somatic anxiety, cognitive anxiety/worry, and concentration-disruption (Dunn et al.,
2000).
The “Competitive State Anxiety lnventory-2” (CSAI-2; Martens et al.,
1990)The “Competitive State Anxiety Inventory-2” (CSAI-2; Martens et al., 1990) is a test of
state anxiety. It comprises twenty-seven items which are divided into three sub-scales
(with each containing nine items): cognitive anxiety, somatic anxiety and self-confidence.
Typical items in the somatic anxiety sub-scale include “I feel nervous” and “My body
feels tense”. A sample item in the cognitive anxiety sub-scale is “I am concerned about
losing”. The “self-confidence” sub-scale is included in the test because a lack of
confidence is believed to be a sign of cognitive anxiety (ibid.). On a four-point scale
(with 1=“not at all” and 4=“very much so”), respondents are required to rate the intensity
of their anxiety experiences prior to competition. Following a review of forty-nine studies
using the CSAI-2, Burton (1998) reported that internal consistency estimates for these
three sub-scales ranged from 0.76 to 0.91.
In the previous section of the chapter, we indicated the importance of athletes’
interpretations of their arousal symptoms. In this regard, the CSAI-2 is hampered by a
significant methodological deficiency—namely, its neglect of the issue of “direction” or
personal meaning of anxiety symptoms for athletes (G.Jones, 1995). To rectify this
problem, some researchers advocate the addition of a directional measure to all
“intensity” indices of anxiety (G.Jones and Swain, 1992). In this case, respondents may
be required first to complete the CSAI-2 in order to elicit the intensity with which they
experience the twenty-seven symptoms listed in this test. Then, they may be asked to rate
the degree to which the experienced intensity of each symptom is facilitative or
debilitative of their subsequent athletic performance. A seven-item Likert response scale
is used, with values ranging from −3 (indicating “very negative”) to +3 (indicating “very
positive”). To illustrate, an athlete might respond with a maximum “4” to the statement “I
am concerned about losing” but might then rate this concern with a +3 on the
interpretation scale. Through these scores, the performer is indicating that s/he feels that
this concern about losing is likely to have a facilitative effect on his/her forthcoming
performance. With this modification, CSAI-2 “direction of anxiety” scores can vary
between −27 and +27. Internal consistency reliability estimates for this
facilitative/debilitative measure range from 0.72 (for the somatic anxiety sub-scale) to
0.83 (for the cognitive anxiety sub-scale) (Swain and Jones, 1996). When this
“directional modification” scale has been used in conjunction with the CSAI-2, the
resulting instrument is called the “DMCSAI-2” (Burton, 1998) or the CSAI-2 (d) (M. V.
Jones and Uphill, 2003). But how valid is this procedure? See Box 3.3.
"Psyching up" and "calming down": anxiety in sport 77