possible range of 0 to120, and participants are supposed to answer the items in 60 minutes.
According to Abedi (2002), the estimated correlation coefficient between the four subscales of the ACT and the Torrance Test of Creative Thinking (TTCT) was meaningful at the 0.01 level of significance. Therefore, the ACT has concurrent validity. It is worth noting that the TTCT is one of the best known tests of creativity (Marashi & Dadari, 2012), which consists of two sections: verbal and nonverbal or figural section. In fact, the development of the ACT was an attempt to shorten the time needed for the administration and scoring of the TTCT (Marashi & Dadari, 2012).
The estimated reliability of each of the subscales of the ACT was 0.61 to 0.75, which demonstrated that the test is also reliable (Auzmendi, Villa, & Abedi, 1996).
ACT is translated by Daemi and Moghimi (2004) with a reliability of 0.595 using the Cronbach’s alpha coefficient. In another study carried out by Marashi and Dadari (2012) the reliability of the ACT was reported as 0.71 using the Cronbach’s alpha coefficent.
The ACT’s translated version of Daemi and Moghimi (2004) was validated by Zaker (2013). Zaker (ibid) followed the following sections to assess and appraise the validity of the instrument:
• Reporting on content validity;
• Checking the criterion-related validity;
• Analyzing of the internal structure of the instrument employing exploratory factor analysis through principal components analysis.
He has also presented a report on the reliability of the instrument. As reported by Zaker (ibid) ten non functioning items were excluded from the questionnaire because they either did not load on any of the factors or loaded simultaneously on a couple of factors.
The 50 multiple-choice items of this questionnaire have three options (Appendix C) ranging from least to most creative responses with a range of scores between 0-2; therefore, the scores of the Persian Creativity Test could range from 0 to 100, and the participants are allocated 50 minutes to respond to the questionnaire.
Table 3.3 demonstrates the seven subcomponents of the Persian Creativity Test and the related items, as reported by Zaker (2013):
Table 3.3: Subcomponents and Items of the Persian Creativity Test
Subcomponents of the Persian ACTa
Number of Items
38, 16, 47,17, 15, 19, 18, 46, 14, 37, 39, 25, 10, 24, 31, 32, 26, 11, 23,33, 2, 6, 7, 8, 1
22, 27, 40,21, 41, 45, 12
43, 44, 50, 42, 48
49, 4, 9, 28
20, 36, 5
34, 35, 3, 30
a. Labeled according to the content of the items with highest eigenvalue in each category
According to Zaker (ibid), the internal consistency of the Persian Creativity Test was estimated to be 0.85 employing Cronbach’s alpha coefficient.
In this study, to avoid any misinterpretation, the Persian version of Creativity Test which has been translated by Daemi and Moghimi (2004) and validated by Zaker (2013) was used. The reliability of creativity questionnaire was estimated to be 0.90 using the Cronbach’s alpha coefficient, which demonstrated a fair degree of reliability. Table 3.4 shows the reliability of creativity questionnaire.
Table 3.4: Reliability Index of Creativity Questionnaire
N of Items
To achieve the purpose of this study and address the questions posed, certain procedures were pursued which are explained hereunder:
A package of three distinct questionnaires were given in one session to both male and female sophomore, and senior students majoring in English Translation and English Literature at Islamic Azad University, Central Tehran in order to collect the required data of this study.
It should be mentioned that at the beginning of the administration procedure, as the researcher was administering the English version of the questionnaires, it was found that the participants did not answer many of the questions. In order to avoid any misunderstanding, because of the lack of vocabulary, grammar knowledge, and cultural differences and also to collect reliable data, the researcher administered the Persian version of the questionnaires.
Before administering the Persian version of questionnaires, the participants were fully briefed on the process of completing the questionnaires; this briefing was given in Persian through explaining and exemplifying the process of choosing answers. The researcher intentionally randomized the order of administered questionnaires to control for the impact of order upon the completion process and validity of the data.
Then, the three questionnaires were administered to 250 participants. The researcher randomly observed the process of filling out for some individuals to make sure they were capable to fully understand the questions and responses. It should be added that the whole length of the class periods of 100 minutes were devoted to administering these questionnaires. Moreover, at administration process of the questionnaires, the responses were handled with confidentiality.
It should be mentioned that from the initial 250 administered questionnaires, a number of 179 were returned to the researcher. Out of that number, 148 sets answered for all the three questionnaires, were considered for statistical analyses regarding the relationship among the variables.
Subsequently, the administrated questionnaires were scored and the data, which was gathered through the above-mentioned procedures, was analyzed by the use of the Statistical Package for Social Sciences (SPSS).
The design of the present study is descriptive as well as ex-post-facto since the primary motivation of the researcher was to investigate the relationship among the three variables with no preceding manipulation (Best & Kahn, 2006; Springer, 2010; Tabachnick & Fidell, 2007). Learning styles and language learning strategies were the predictor variables of the predicted variable, creativity. Furthermore, age and gender were categorized as intervening variables.
3.6 Statistical Analyses
In the present study, the data analysis provided descriptive statistics and inferential statistics as well. Descriptive statistics like mean, standard deviation, and standard error of the mean were obtained. The reliability of the research instruments were estimated through Cronbach’s alpha coefficient. Afterwards, to check the normality of distribution, skewness ratio and kurtosis ratio were calculated. Moreover, assumptions of linear correlations were checked to decide whether parametric or non-parametric techniques should be employed.
Given the inferential statistics, considering the fact that the assumptions of linear correlation were met, the Pearson’s product-moment correlation as a parametric technique was used in order to investigate the relationship among EFL learners’ learning style preferences, language learning strategies and creativity. To verify the four primary research hypotheses, the scores on PLSP, SILL, and ACT were correlated, and in order to address the fifth, sixth, seventh, and the eighth questions, the multiple regression analyses were employed. Details of the above-mentioned analyses are appeared in the fourth chapter.
RESULTS AND DISCUSSION
In the process of this study, the researcher conducted a series of pertinent calculations and statistical routines in order to test the hypotheses formulated and came up with certain results that are elaborated comprehensively in this chapter. All the data analysis procedures and results are presented and discussed in the order of checking the assumptions of linear correlation, testing the research hypotheses, and reporting the reliability of the instruments.
As mentioned before, the purpose of this study was to see whether any significant relationship exists among EFL learners’ use of language learning strategies, learning style preferences, and creativity. Therefore, the following hypotheses were stated:
H 01: There is no significant relationship among EFL learners’ use of language learning strategies, learning style preferences, and creativity.
H02: There is no significant relationship between using different types of language learning strategies and learning style preferences by EFL learners.
H 03: There is no significant relationship between EFL learners’ use of language learning strategies and their creativity.
H 04: There is no significant relationship between EFL learners’ learning style preferences and their creativity.
Provided that a significant correlation is obtained for the variables, the following hypotheses were also raised:
H05: There is no significant difference among EFL learners’ use of language learning strategies, and learning style preferences, in predicting creativity.
H 06: EFL learners’ learning style preferences do not predict their use of language learning strategies.
H 07: EFL learners’ use of language learning strategies does not predict their creativity.
H 08: EFL learners’ learning style preferences do not predict their creativity.
4.2 Results of the Study
The randomly selected, 250 participants were given the three questionnaires, from which 179 questionnaires were returned to the researcher. Out of 179 questionnaires, 148 series which were filled out completely used for further statistical analyses. In order to test the null hypotheses of the study, the descriptive statistics were obtained and the assumptions of linear correlation were checked, the results of which are presented hereunder.
4.2.1 Testing Assumptions
Five assumptions should be met before one decides to run Pearson correlation or regression to analyze