2.1 Sample
The pilot sample of the study involved 102 university students [58 (57%) women, 44 (43%) men] from different Departments of Greek Universities, who completed a set of online self-report questionnaires. The pilot study did not indicate the need to modify the questionnaires. Therefore, the pilot sample (N = 102) was integrated into the sample of the main study (N = 690 students) resulting in a total sample of 792 university students [380 (48%) women, 412 (52%) men]. Out of them, 59% were undergraduate and 41% were postgraduate (MSc, PhD, Postdoc) students. Most of the undergraduate students were in the second year of their studies (70%), while most of the postgraduate students (85%) attended a Master (MSc) program. Due to this unequal distribution, we were not statistically allowed to examine the effect of the year of undergraduate studies, or the effect of the postgraduate program attended (MSc, PhD, Postdoc) on the variables under study. As a result, the sample is presented in two broad levels of studies (undergraduate and postgraduate). Regarding the Department of studies, students studied in Departments of Schools of Humanities and Social (37%), Polytechnic (30%), and Economic Sciences (33%). Finally, regarding the age of the students (Μean = 23, SD = 4.70), 120 (42%) were 20 years old, 102 (35%) were 24 years old, 30 (10%) were 21 years old, 15 (5%) were 28 years old, 12 (4%) were 22 years old, 10 (3%) were 19 years old, and 3 (1%) were 34 years old.
2.2 Instruments
For the purpose of the present study, a set of online self-reported questionnaires was used. After the introductory demographic questions (e.g., gender, age, and level/Department of studies), four main parts followed:
Facebook Intensity Use Scale: Students’ perceived intensity of Facebook use was measured with the Greek translation of the Facebook Intensity Use Scale (Ellison et al.
2007), which has been previously used in a sample of Greek adolescent students with good psychometric properties (
α = .83) (Dimogiorga and Syngollitou
2015). This scale was selected due to the fact that it is the only one in international literature that evaluates the use of social networking sites, such as Facebook, including questions and statements that measure not only the frequency of its use but also the extent of students’ active involvement in the activities of Facebook. The scale evaluates: (a) the number of Facebook friends based on an eight-point scale (from 1 =
up to 10 friends to 8 =
over 400 friends) and (b) the time spent on Facebook per day based on a five-point scale (from 1 =
up to 10 min to 5 =
over 3 h). Furthermore, the scale includes six statements to tap the extent to which participants are emotionally connected to Facebook and the extent to which Facebook is integrated into their daily activities (e.g., “Facebook is part of my daily routine activity”, “I am proud when I tell others that I have Facebook profile too”), reflecting in that way the intensity of Facebook use. Participants are asked to respond to the statements based on a five-point Likert scale, ranging from (1) =
not valid at all to (5) =
absolutely valid. According to the initial constructors of the scale (Ellison et al.
2007), intensity of Facebook use can be reflected not only in a qualitative way, such as an individual’s emotional connectedness to Facebook activities (e.g., chatting and posting), but through quantitative dimensions as well, such as the number of Facebook friends and the time spent on Facebook daily. In this way, the scale tries to reflect a multidimensional and more generalized sense of an individual’s intensity of Facebook use. The index of Facebook Intensity Use derives from the average of the total score of the questions and the statements, as long as they have been previously converted to standardized
z scores due to the differing item scale ranges. Based on this standardization, the cutoff scores range from 8 to 48 points with 24 points (Mean = 2.4, SD = 1.21) corresponding to an average intensity of Facebook use. The higher the score (and the corresponding Mean) the higher the intensity of Facebook use (see Ellison et al.
2007).
In order to test the factorial validity of the scale in the university student sample, a principal component analysis was carried out using the main component method and Varimax-type rotation (KMO = .892, Bartlett Chi-square = 1982.133,
p < .001). One factor emerged with eigenvalue > 1.0 and significant interpretive value (Table
1), confirming the unidimensional structure of the scale (Dimogiorga and Syngollitou
2015; Ellison et al.
2007): Factor 1 = Intensity of Facebook use, explaining 51.22% of the total variance. The internal consistency index for the factor is
α = .823. The affinities (according to Pearson’s correlation coefficient r) of the score of each question of the factor with the sum of the scores of the remaining questions of the factor (corrected item—total correlation) are considered satisfactory [in a sample of 300 and 600 people, loadings of more than .29 and .21, accordingly, are accepted (Field
2005)]: Factor 1 (from
r = .47 to
r = .83).
Table 1
Principal component analysis of the Facebook intensity use scale (N = 792)
1. About how many total Facebook friends do you have? | .488 |
2. On average, approximately how many minutes per day do you spent on Facebook? | .512 |
3. Facebook is part of my everyday activity | .733 |
4. I am proud to tell people I’m on Facebook | .551 |
5. Facebook has become part of my daily routine | .872 |
6. I feel out of touch when I haven’t logged onto Facebook for a while | .719 |
7. I feel I am part of the Facebook community | .698 |
8. I would be sorry if Facebook shut down | .544 |
Self-Disclosure Index: Students’ self-disclosure to Facebook friends met on Facebook (unknown online friends) was examined with the Greek translation of the Self-disclosure Index (Miller et al.
1983), which has been previously used in a sample of Greek adolescent students with good psychometric properties (
α = .92) (Dimogiorga and Syngollitou
2015). The “Self-Disclosure Index”, which includes ten statements, is the only one in the related literature measuring the frequency at which a person reveals personal information on various subjects (e.g., “I reveal my personal interests and hobbies”, “I reveal my deepest feelings”) on Facebook friends met on Facebook (unknown online friends). Responses are given on a four-point Likert scale (from 1 =
never to 4 =
often). Individual items are summed to produce an overall score ranging from 10 to 40 points with 20 points (Mean = 2.0, SD = 1.09) corresponding to an average level of self-disclosure to unknown Facebook friends. Therefore, higher scores (and corresponding Means) indicate higher levels of self-disclosure on Facebook (see Miller et al.
1983).
In order to test the factorial validity of the Index in the university student sample, a principal component analysis was carried out using the main component method and Varimax-type rotation (KMO = .845, Bartlett Chi-square = 1911.409,
p < .001). One distinct factor emerged with eigenvalue > 1.0 and significant interpretive value (Table
2) in-line with the original unidimensional structure, confirming the unidimensional structure (Dimogiorga and Syngollitou
2015; Miller et al.
1983): Factor 1 = Self-Disclosure, explaining 51.98% of the total variance. The internal consistency indexes for Factor 1 is
α = .85. The affinities (according to Pearson’s correlation coefficient r) of the score of each question by Factor 1 with the sum of the scores of the remaining questions of the factor (corrected item—total correlation) are considered satisfactory: Factor 1 (from
r = .39 to
r = 59).
Table 2
Principal component analysis of the self-disclosure index scale (N = 792)
1. I reveal my personal interests and hobbies | .841 |
2. I reveal things I've done for which I feel guilty | .772 |
3. I reveal things I wouldn't do in front of people | .689 |
4. I reveal my deepest feelings | .754 |
5. I reveal what I like and what I don't like about myself | .791 |
6. I reveal what's important to me in life | .819 |
7. I reveal what makes me who I am | .858 |
8. I reveal my worst fear | .701 |
9. I reveal things I've done, and I feel proud of | .599 |
10. I reveal my close relationships with other people | .801 |
Resilience Scale: Students’ resilience was examined with the Greek translation of the short version of the Connor-Davidson Resilience Scale (The Connor-Davidson Resilience Scale—CD-RISC; Connor and Davidson
2003) of Campbell-Sills and Stein (
2007). The original long version of the CD-RISC (25 items) investigates individuals’ positive adaptation to stressful and/or difficult situations. According to Campbell-Sills and Stein (
2007), the factor structure of the original CD-RISC across demographically equivalent samples is unstable, while the short version of the scale has good psychometric properties (
α = .89) in a sample of undergraduate university students. This finding demonstrates that resilience can be reliably assessed with a subset of the CD-RISC items. Thus, in the short version of the CD-RISC, resilience is measured through ten representative statements/proposals (they reflect individuals’ ability to tolerate experiences such as change, personal problems, illness, pressure, failure, and painful feelings) which form the single factor “resilience” (Campbell-Sills and Stein
2007). These statements/proposals are answered on a 5-point Likert scale (from 0 =
not at all true to 4 =
almost always true). Examples of the statements/proposals are the following: “I tend to bounce back after illness or hardship”, “I can stay focused under pressure”, and “I am able to adapt to change”. Individual items are summed to produce an overall score ranging from 0 to 40 points with 20 points (Mean = 2.0, SD = 1.91) corresponding to an average level of resilience. Therefore, higher scores (and corresponding Means) indicate higher levels of resilience (see Campbell-Sills and Stein
2007).
In order to test the factorial validity of the scale, a principal component analysis was carried out using the main component method and Varimax-type rotation (KMO = .833, Bartlett Chi-square = 1499.311,
p < .001). One distinct factor emerged with eigenvalue > 1.0 and significant interpretive value (Table
3) in-line with the original factor structure: Factor 1 = Resilience, explaining 44.12% of the total variance. The internal consistency indexes for Factor 1 is
α = .811. The affinities (according to Pearson’s correlation coefficient
r) of the score of each question by Factor 1 with the sum of the scores of the remaining questions of the factor (corrected item—total correlation) are considered satisfactory: Factor 1 (from
r = .41 to
r = 88).
Table 3
Principal component analysis of the resilience scale (N = 792)
1. I am able to adapt to change | .502 |
4. I can deal with whatever comes | .599 |
6. I try to see humorous side of problems | .702 |
7. Coping with stress can strengthen me | .689 |
8. I tend to bounce back after illness or hardship | .581 |
11. I can achieve goals despite obstacles | .699 |
14. I can stay focused under pressure | .505 |
16. I am not easily discouraged by failure | .791 |
17. I thinks of self as strong person | .499 |
19. I can handle unpleasant feelings | .551 |
Loneliness Scale: Students’ sense of loneliness was examined with the Greek translation of the short form of California Los Angeles Loneliness Scale—UCLA-LS (ULS-8; Hays and DiMatteo
1987), which has been previously used in samples of university students with satisfactory psychometric properties (
α = .72) (Doğan et al.
2011; Wu and Yao
2008). According to previous studies, there are concerns with the factorial validity of the original 20-item UCLA-LS, indicating that a short form is also reliable and valid as the original 20-item scale while also displaying superior model fit and reduce the burden on respondents (Doğan et al.
2011; Wu and Yao
2008). This scale includes eight proposals/statements, for which individuals are asked to state how often they feel as each proposal/statement describes on a four-point Likert scale ranging from (1) =
never to (4) =
always. Specifically, the scale consists of two positively worded (“I am an outgoing person,” and “I can find companionship when I want it”), which are reverse scored. Individual items are summed to produce an overall score ranging from 20 to 80 points. Based on these cutoff scores and considering the four-point Likert scale 40 points correspond to an average sense of loneliness (Mean = 2.0, SD = 1.09). The higher the total score of the scale (and the corresponding Mean) the higher the sense of loneliness (Hays and DiMatteo
1987). Examples of the proposals/statements are the following: “People are around me but not with me”, “I lack companionship”, and “There is no one I can turn to”.
In order to test the factorial validity of the scale, a principal component analysis was carried out using the main component method and Varimax-type rotation (KMO = .872, Bartlett Chi-square = 1443.144,
p < .001). One factor emerged with eigenvalue > 1.0 and significant interpretive value (Table
4), confirming the unidimensional structure of Hays and DiMatteo (
1987): Factor 1 = Loneliness, explaining 32.11% of the total variance. The internal consistency indexes for this Factor is
α = .821. The affinities (according to Pearson’s correlation coefficient
r) of the score of each question by the factor with the sum of the scores of the remaining questions of the factor (corrected item—total correlation) are considered satisfactory: from
r = .48 to
r = .58.
Table 4
Principal component analysis of the loneliness scale (N = 792)
1. I lack companionship | .598 |
2. There is no one I can turn to | .711 |
3. I am an outgoing persona | .742 |
4. I feel left out | .528 |
5. I feel isolation from others | .633 |
6. I can find companionship when I want ita | .701 |
7. I am unhappy being so withdrawn | .577 |
8. People are around me but not with me | .723 |