2.1. Implementing the Survey on Moodle
The survey was conducted over two years (2018–2019) for both full-time students at Lappeenranta-Lahti University of Technology (LUT) University and distance learning students who study alongside their work. LUT University (Lappeenranta-Lahti University of Technology (LUT)) is a science university in Finland, bringing together the fields of science and business. The questionnaire was targeted at both full-time and distance learning students in the Bioenergy course, who were given the opportunity to answer questions electronically at the end of the course. Responses were not given anonymously, and respondents were aware of this. Full-time students have regular face-to-face lectures, while distance learners self-study through the digital platform Moodle. Both groups had the same lecture material available; only the teaching method was different. Distance learning students also had the opportunity to ask questions of the teacher through a digital platform. The survey was conducted in a digital learning environment. MoodleTM, “Modular Object-Oriented Dynamic Learning Environment”, is a free, open-source object-based learning platform, i.e., virtual learning environment (VLE). Moodle can be used to build courses that allow you to publish material (e.g., in a timed fashion) and conduct experimental tests. Moodle provides tools for interaction, content production and material sharing, among other things. The Bioenergy course was conducted in this environment for both student groups. One of the topics of the course was the sustainability of bioenergy supply and use. All dimensions of sustainability, such as environmental, social and economical, were included in the definition of sustainability. In a digital learning environment, each group of students was able to answer the question option in their own time and justify their answer. Answers were not graded and there was no length limit to the justification of the answers. The choice question offered three options, “yes”, “no” and “don’t know”, from which the respondent chose one.
The questions were:
There was no attempt to guide the students in their answers, because there could be reasons to be either positive, negative or not know opinion. In practice, the course material and the way it was delivered may affect responses compared to a hypothetical reference group that would not have seen the teaching material. Therefore, the group of students taking the course are not representative of a broader population of students that didn’t just learn about sustainability and bioenergy. However, students were encouraged to think critically and source criticism. The course does not tell whether the use of forest-based bioenergy is sustainable or not but highlights various methodological choices which may influence to the outcome. Those are, e.g., definition of reference (no bioenergy) scenario, time frame of evaluation period, spatial scale, scope (like one product life cycle or system level assessment) and metric choice. The survey gave an idea of how these different groups view the use of forest-based bioenergy and how different options were justified.
The number of distance learning students was larger than the number of full-time students each year, and the number of students increased in 2019 (Table 1
). The response rate for full-time students was 70% and 100% for distance learners who completed the query as part of the course. A total of 273 students participated in the study. For full-time students, answering was voluntary, resulting in a lower response rate.
Background information such as gender, nationality and field of study was collected from students (Figure 2
and Figure 3
). Nationality was classified into groups of Finnish and foreigners because the group of foreigners was divided into so many nationalities. The background information was obtained directly from the system and did not need to be requested separately. Full-time students could also be graded according to the progress of their studies, either bachelor’s (BSc) or master’s (MSc). All distance learning students were master’s students. The group of distance learning students clearly differed from the group of full-time students. It was more male dominated, the majority were Finns, and the study fields were more concentrated. The group of distance learning students were working people who had a bachelor’s degree and were generally older and more experienced than full-time students. However, age was not collected separately, but there was a difference in age between groups.
2.2. Testing the Background Variables
Responses were cross tabulated for background variables. This was to determine whether the background variable had an effect on the response. In cross tabulation, observations were presented in number distributions and a hypothetical tabulation with no dependence on background variables was constructed. The numbers in the hypothetical table are called expected frequencies. In order to make the frequencies high enough (>5), the answers “no” and “don’t know” were combined, as were the individual observations in the field of studies, under the group “other”. Testing the difference between the observed table and the hypothetical table is not reliable if the numbers of the hypothetical table, i.e., the expected frequencies, are too small as in the groups “don’t know” and “no”. “Don’t know” is not a same as “no” answer, but they were expected to be more sceptical than those who answered “yes”. However, the justification for all response options was further analysed. The justification was categorised under certain keywords which describe the answers in the best possible way, but students were not encouraged to use any keywords.
The difference between the observed and the hypothetical table was measured by the Chi-square test variable (X2). The Chi-square test variable is known to follow a roughly Chi-square probability distribution, whose exact shape depends on the degree of freedom (df). The degrees of freedom are calculated from the table (number of rows−1) × (number of columns−1). The value of the Chi-square test variable is higher the more the observed frequencies deviate from the expected frequencies.