Oldal 95 [95]
STUDY Two
In the last part of the questionnaire, a series of questions elicited informa¬
tion about the background ofthe participants. The questions concerned gender,
age, number of years learning English, type of secondary school, number of
EFL lessons a week, whether participants have any foreign language exam
certificates, and whether participants took part in foreign exchange programs
before.
5.3 DATA COLLECTION AND THE VALIDATION PROCESS
The questionnaire was developed in Hungarian, and the adopted English lan¬
guage scales (where necessary) were translated into Hungarian. Through expert
judgment, two professionals and two potential participants (secondary school
students) were asked to complete the questionnaire implementing the so-called
think-aloud protocol, which meant that these people were asked to think aloud
while completing the questionnaire. This was important to make sure that all
questions and items were interpreted the same way and the way they were
meant to be interpreted. All problematic questions were reworded and fine¬
tuned before the pilot process. The questionnaire was piloted with 50 students,
and in order to ensure the internal consistency of the scales, Cronbach's alpha
coefficients were calculated for all scales. Based on Cronbach’s alpha coefficients,
some of the items were fine-tuned and reworded once more so that no am¬
biguous items could remain in the questionnaire (for more information on the
validation process, see Fajt 2021b). The final version of the questionnaire was
sent to Hungarian secondary school students learning English as a first or
second foreign language at secondary school, and the data collection itself took
place in March 2021.
5.4 DATA ANALYSIS
The obtained data were first coded; then, all negatively worded items (reversed
items) were recoded in reverse. The coded data was then input into SPSS (Sta¬
tistical Package for Social Sciences) 27.0, which allowed for complex statistical
analyses with the statistical significance level set for p<.05. Before the analyses,
as explained previously, Cronbach’s alpha internal consistency coefficients
were calculated for the scales to investigate their reliability. This was followed
by descriptive statistical measures when mean scores (M) and standard de¬
viations (SD) were calculated for all scales. Then the mean scores of scales were
compared through independent samples t-tests in the case of questions with
two answers (e.g., gender). Additionally, paired samples t-tests were also used
to compare different scales measuring extramural and in-class dimensions
+ 93 +