3. Method

#### 3.1. Instrument

The proposed measurement instrument is based around four dimensions and 21 items as stated above. The reliability and validity studies for the instrument used show rates of internal consistency that are more than acceptable or satisfactory (between 0.85 and 0.92, overall reliability) and the exploratory and confirmatory factor analyses indicate the presence of a single construct, based on three correlated dimensions, explaining 42.38% of the variance. The principal components analysis procedure was used as a factor extraction method. The direct Oblimin factor rotation method was used.

The indices analysed to evaluate fit (Chi-squared/degrees of freedom, TLI, CFI and RMSEA) show that this is the model with the best fit.

A detailed study into the reliability and validity of this instrument is provided in García Ramos et al. [17].

#### 3.2. Population, sample and data collection process

The study population comprises second-year students at the UFV who are taking this module. A total of 639 students were enrolled in this module in the 2012–2013 academic year. The sample was selected through a quota sampling process, taking the course as the main characteristic, in which the sample should be similar to the study population, second year students at the UFV.

The information was collected using an individual survey. The field work was performed two times: (1) in October 2012 it was performed on the students when teaching of the module had recently started (pretest) and (2) in May 2013 it was performed a second time when the students had completed the module (posttest). The samples from the pretest and the posttest are independent as the questionnaires are anonymous and the individuals who answered the pretest and posttest are not identified to make comparison of the samples possible. Table 1 shows the technical details of the sampling.

A total of 757 surveys was eventually obtained: 404 surveys in the pretest (63.2% of the population) and 353 surveys in the posttest (55.2% of the population). The following profile of the sample was subsequently identified (Table 2).

#### 3.3. Data analysis process

2. Method, hypothesis and results

impact on the students in each course.

Oblimin factor rotation method was used.

show that this is the model with the best fit.

3.2. Population, sample and data collection process

The general goal of this study is to evaluate the impact of taking the SREU module on students

• To find possible common patterns between courses regarding the factors that comprise

There are three main hypotheses to compare in this study. They relate to three types of possible impact: (1) the module has a general impact on the students; (2) the different dimensions that make up SREU have different impacts on the students; and (3) the module has a different

The proposed measurement instrument is based around four dimensions and 21 items as stated above. The reliability and validity studies for the instrument used show rates of internal consistency that are more than acceptable or satisfactory (between 0.85 and 0.92, overall reliability) and the exploratory and confirmatory factor analyses indicate the presence of a single construct, based on three correlated dimensions, explaining 42.38% of the variance. The principal components analysis procedure was used as a factor extraction method. The direct

The indices analysed to evaluate fit (Chi-squared/degrees of freedom, TLI, CFI and RMSEA)

A detailed study into the reliability and validity of this instrument is provided in García

The study population comprises second-year students at the UFV who are taking this module. A total of 639 students were enrolled in this module in the 2012–2013 academic year. The sample was selected through a quota sampling process, taking the course as the main

from a variety of university courses, using a reliable and valid tool.

• To measure whether different courses have an impact.

2.1. Goals

26 Social Responsibility

The specific goals are:

SREU.

2.2. Hypothesis

3. Method

3.1. Instrument

Ramos et al. [17].

The SPSS version 21 statistics program was used to codify and analyse the data (SPSS Inc., Chicago, USA). Descriptive analyses were performed (central tendency measures and standard deviation). To compare the significant differences at the statistical level between the two moments (before and after taking the module), Student's t-test for independent samples and one factor and two factor Analyses of Variance (ANOVA) parametric tests were used to analyse the differences. The assumptions of independence of the data (normality and homoscedasticity) were tested. Student's t-test was performed, without distinguishing between courses, on an aggregate dimension (the sum of the four dimensions). One-factor analyses of variance (ANOVA) were performed on each of the four dimensions and on the opinion item. For these six quantitative variables, two-factor ANOVA analyses were performed to evaluate the effects of the two possible sources of variance, as well as the interaction between them: (A) belonging to the pretest or posttest groups and (b) belonging to a particular course. One-factor ANOVA analyses were also performed on the four dimensions and the opinion item for each of the courses. A confidence level of 95% or higher was used in almost all of the analyses.


Table 1. Technical details of the research project.

#### 28 Social Responsibility


Table 2. Profile of the sample.
