Interactive effects of English-speaking anxiety and strategy use on oral English test performance of high- and low-proficient Chinese university EFL learners

Figures & data

Table 1. Characteristics of instruments (N = 1092)

	No. of items	Reliability	Mean item-total correlation (p = .01)
ESAS1	10	.81	.298
ESAS2	7	.784	.342
ESAS	17	.882	.305
OCSI1	6	.823	.356
OCSI2	6	.843	.315
OCSI3	5	.783	.278
OCSI4	4	.771	.189
OCSI5	4	.801	.305
OCSI6	3	.79	.301
OCSI7	2	.728	.273
OCSI8	2	.762	.254
OCSI	32	.874	.178

Notes: ESAS1 = fear of negative evaluation; ESA2 = communication apprehension; OCSI1 = social affective strategies; OCSI2 = fluency-oriented strategies; OCSI3 = accuracy-oriented strategies; OCSI5 = message abandonment strategies; OCSI4 = negotiation for meaning while speaking strategies; OCSI6 = message reduction and alteration strategies; OCSI7 = nonverbal strategies while speaking; OCSI8 = attempt to think in English strategies; coefficient of determination: small = r ≤ 0.1; medium = r = 0.3; large = r ≥ 0.5 (Cohen, 1988)

Table 2. Means and standard deviations of ESAS and OCSI scales for low- and high-proficiency groups

	Low-proficiency (N = 178)		High-proficiency (N = 214)		t-test results
	Mean	SD	Mean	SD	t	p	Effect size (Cohen’s d)
ESAS1	2.85	.735	2.53	.582	2.61	.022	.26
ESAS2	3.13	.802	2.71	.564	2.84	.001	.29
OCSI1	3.36	.685	3.77	.647	−2.67	.000	.27
OCSI2	3.33	.791	3.58	.704	−2.38	.031	.24
OCSI3	3.34	.603	3.45	.534	−.106	.795	/
OCSI4	3.19	.781	3.51	.649	−2.91	.001	.29
OCSI5	3.13	.647	2.87	.582	3.33	.001	.34
OCSI6	3.21	.757	3.63	.801	−3.88	.000	.39
OCSI7	3.43	.841	3.13	.794	2.88	.001	.29
OCSI8	3.27	.876	3.03	.761	2.38	.034	.24

Notes: effect size of Cohen’s d: small = d ≤ 0.2; medium = d = 0.5; large = d ≥ 0.8 (Cohen, 1988)

Table 3. Correlations between ESAS and OCSI scales

	Low-proficient group (N = 178)		High-proficient group (N = 214)
	ESAS1	ESAS 2	ESAS 1	ESAS 2
OCSI1	−.331** .000	−.456** .000	−.320* .000	−.300** .000
OCSI2	−.421** .000	−.481** .000	−.598** .000	−.641** .000
OCSI3	−.280** .000	−.403** .000	−.427** .000	−.517** .000
OCSI4	−.341** .000	−.351** .000	−.342** .000	−.247** .000
OCSI5	.426** .000	.338** .000	.164 .014	.183 .012
OCSI6	.242* .002	.197 .018	−.345** .000	−.161 .014
OCSI7	−.516** .000	−.552** .000	−.495** .000	−.372** .000
OCSI8	−.006 .839	−.019 .836	.011 .871	−.173 .013

Notes: * = p ≤ .005;

** = p ≤ .001; coefficient of determination: small = r ≤ 0.1; medium = r = 0.3; large = r ≥ 0.5 (Cohen, 1988)

Table 4. Regression coefficients and significance of predictors for test performance

		Speaking test performance
		β	t	p	VIF	Variance	Cohen’s f^2
Low-proficient group (N = 178)	ESAS1	−.282	−2.33^**	.003	1.000	7.2%	.031
	OCSI2	.221	2.29^**	.034	1.000	5.1%	.034
	OCSI5	−.137	−2.17^**	.037	1.000	2.1%	.076
High-proficient group (N = 214)	OCSI2	.256	2.35^**	.011	1.000	4.6%	.021
	ESAS2	−.145	−2.21^**	.023	1.000	2.3%	.087

Notes: ** = p ≤ .01; effect size of Cohen’s f²: small = f² ≤ .02; medium = f² = .15; large = f² ≥ .35 (Cohen, 1988)

Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). New Jersey: Lawrence Erlbaum Associates.

 Google Scholar