Sample characteristics
The online questionnaire was completed by a total of 257 students at the University for Continuing Education Krems (UWK) and by a total of 96 students at Johannes Kepler University Linz (JKU). Table
1 indicates the sample characteristics of the participating UWK and JKU students. In the following, we refer to the participants in our study as “JKU students” or “UWK students” for the sake of clear distinction. Neither term implies that the study allows to derive generalizable statements for the whole population of either university—the present study focusses on comparing the differences between the student groups, which pursue their studies under different framework conditions, rather than comparing the two involved universities in particular.
Table 1
Sample demographics and household characteristics of UWK and JKU students
Gendera (df = 1, n = 349) |
Male | 107 | 41.6 | 57 | 59.4 | 8.153** |
Female | 146 | 56.8 | 39 | 40.6 |
Diverse | 1 | 0.4 | – | – |
No indication made | 3 | 1.2 | – | – |
Agea (df = 3, n = 351) |
< 21 years | – | – | 6 | 6.3 | 226.019** |
21–25 years | 9 | 3.5 | 69 | 71.9 |
26–30 years | 31 | 12.1 | 13 | 13.5 |
> 30 years | 215 | 83.7 | 8 | 8.3 |
No indication made | 2 | 0.8 | – | – |
Experience with online learninga (df = 1, n = 353) |
No experience before | 140 | 54.5 | 51 | 53.1 | .051 |
Experience before | 117 | 45.5 | 45 | 46.9 |
Household forma (df = 3, n = 349) |
Multi-person household | 199 | 77.4 | 87 | 90.6 | 32.660** |
One-person household | 50 | 19.5 | – | – |
Shared apartment | 4 | 1.6 | 7 | 7.3 |
Other household formc | – | – | 2 | 2.1 |
No indication made | 4 | 1.6 | – | – |
Household structureb (df = 1, n = 347) |
No children | 136 | 52.9 | 57 | 59.4 | 1.316 |
Children of compulsory school age | 54 | 21.0 | 13 | 13.5 | 2.484 |
Children of preschool age | 38 | 14.8 | 8 | 8.3 | 2.525 |
Children no longer of compulsory school age | 23 | 8.9 | 7 | 7.3 | .235 |
Household with more than two generations | 16 | 6.2 | 12 | 12.5 | 3.834 |
Household with pets | 77 | 30.0 | 24 | 25.0 | .798 |
No indication made | 5 | 1.9 | 3 | 3.1 | .449 |
Living environmenta (df = 3, n = 351) |
Urban | 85 | 33.1 | 41 | 42.7 | 3.248 |
Suburban | 63 | 24.5 | 23 | 24.0 |
Village environment | 72 | 28.0 | 23 | 24.0 |
Rural | 35 | 13.6 | 9 | 9.4 |
No indication made | 2 | 0.8 | – | – |
Residential building typea (df = 3, n = 335) |
Detached single-family house | 108 | 42 | 36 | 37.5 | 21.230** |
Semi-detached or terraced house | 19 | 7.4 | 7 | 7.3 |
Multi-party house | 114 | 44.4 | 43 | 44.8 |
Other residential building typed | – | – | 8 | 8.3 |
No indication made | 16 | 6.2 | 2 | 2.1 |
Flat sizea (df = 3, n = 353) |
< 40 m2 | 15 | 5.8 | 31 | 32.3 | 45.486** |
40–70 m2 | 53 | 20.6 | 17 | 17.7 |
70–120 m2 | 91 | 35.4 | 29 | 30.2 |
> 120 m2 | 98 | 38.1 | 19 | 19.8 |
Access to outdoor spacesb (df = 1, n = 353) |
Garden | 139 | 54.1 | 42 | 43.8 | 2.988 |
Terrace | 92 | 35.8 | 30 | 31.3 | .639 |
Loggia | 106 | 41.2 | 38 | 39.6 | .080 |
No access to outdoor space | 37 | 14.4 | 27 | 28.1 | 8.874** |
Majority of UWK students are female (56.8%), the majority of JKU students are male (59.4%). A Chi2-Test showed a significant difference in the genders’ distribution between UWK and JKU, X2 (1, 349) = 8.153, p = 0.004. Further significant differences between UWK and JKU were found for the following characteristics: age, X2 (3, 351) = 226.019, p < 0.001; household form, X2 (1, 349) = 32.660, p < 0.001; residential building type, X2 (1, 335) = 21.230, p < 0.001; flat size, X2 (1, 353) = 45.486, p < 0.001, no access to outdoor space X2 (1, 353) = 8.874, p = 0.003.
Perception of home learning environments for digitally supported learning
We examined how students from the two institutions (UWK and JKU) perceived the adequacy of their predominantly used learning environment’s physical and spatial characteristics. We therefore asked students to which extent they agreed that their personal requirements were met at their learning place regarding 11 different attributes on spatial characteristics and indoor environmental conditions. An exploratory factor analysis (EFA) with extraction method of principal component analysis and varimax rotation with Kaiser normalization was used to discover underlying factors. The Kaiser–Meyer–Olkin (KMO) measure of sampling adequacy was 0.843, and Bartlett’s Test of Sphericity was statistically significant (χ2(55) = 1343.55,
p < 0.001). The EFA produced a two-factor solution (eigenvalues λ > 1), explaining 52.90% of the variance. The results of the factor analysis with factor loadings are presented in Table
3.
Table 3
Exploratory factor analysis (EFA) on physical–spatial attributes of home learning environments
08 distraction-free environment | 0.78 | 0.14 | 0.63 |
02 ergonomic work-compatible furniture | 0.77 | 0.02 | 0.59 |
09 protection against noise pollution | 0.70 | 0.30 | 0.57 |
11 adaptability to individual spatial requirements | 0.63 | 0.39 | 0.55 |
01 adequate size | 0.60 | 0.41 | 0.52 |
03 appropriate technical equipment | 0.56 | 0.21 | 0.36 |
07 good ventilation conditions | 0.19 | 0.76 | 0.61 |
04 adequate supply of daylight | 0.10 | 0.69 | 0.49 |
06 comfortable temperature conditions | 0.29 | 0.68 | 0.55 |
05 pleasant view | 0.13 | 0.64 | 0.43 |
10 attractive interior design | 0.35 | 0.63 | 0.52 |
Eigenvalues λ | 3.02 | 2.80 | 5.82 |
Variance explained (%) | 27.5 | 25.4 | 52.9 |
The two factors were interpreted as follows:
Factor 1 Learning place quality (explained 27.5% of variance).
Factor 2 Indoor environmental quality (IEQ) (explained 25.4% of variance).
The variance in students’ perceptions of the physical learning environment is mainly explained by the "Learning Place Quality" and the "Indoor Environmental Quality".
Internal consistency for each of the scales was examined using Cronbach’s alpha. The alphas were acceptable: 0.81 for Learning Place Quality (6 items) and 0.73 Indoor Environmental Quality (5 items). Table
4 presents the results of the reliability analysis. Furthermore, for the items of the factors, the item selectivity (
rit) and mean values (
M) as well as standard deviations (
SD) were calculated.
Table 4
Reliability analysis on F1 and F2
F1 Learning Place Quality | 6 | 340 | 0.81 | 0.46 – 0.64 | 3.13 (0.87) |
F2 Indoor Environmental Quality | 5 | 351 | 0.73 | 0.48 – 0.57 | 3.48 (0.54) |
Independent t-tests (Table
5) indicated a statistically significant difference between students of UWK and students of JKU regarding F2 ‘Indoor Environmental Quality’.
Table 5
Perceptions on learning place quality and indoor environment quality by university
F1 Learning Place Quality | 256 | 3.15 (0.68) | 95 | 3.08 (0.61) | 0.93 (349), 0.354 | – |
F2 Indoor Environmental Quality | 257 | 3.53 (0.51) | 96 | 3.36 (0.48) | 2.84 (351), 0.005** | 0.34 |
Differences between students of UWK and students of JKU regarding the attributes of F1 ‘Learning Place Quality’ and F2 ‘Indoor Environmental Quality’ were analysed with Mann–Whitney U tests. UWK students had statistically significantly better perceptions of their home learning environment in terms of size, protection against noise pollution, good ventilation conditions, attractive interior design, and pleasant view (cf. Table
6), while JKU students reported statistically significantly better perceptions on the availability of ergonomic work-compatible furniture.
Table 6
Different aspects of learning environment by university
| F1 learning place quality |
03 appropriate technical equipment | 256 | 3.52 (0.72) | 4.00 | 96 | 3.52 (0.67) | 4.00 | 12,113.00 | − 0.240 | 0.810 | – |
01 adequate size | 256 | 3.42 (0.86) | 4.00 | 96 | 3.22 (0.89) | 3.00 | 10,542.50 | − 2.325 | 0.020* | 0.22 |
09 protection against noise pollution | 255 | 3.15 (0.94) | 3.00 | 95 | 2.82 (0.99) | 3.00 | 9828.00 | − 2.875 | 0.004** | 0.29 |
11 adaptability to individual spatial requirements | 252 | 3.15 (0.94) | 3.00 | 92 | 3.01 (0.94) | 3.00 | 10,572.50 | − 1.332 | 0.183 | – |
08 distraction-free environment | 257 | 2.99 (1.03) | 3.00 | 96 | 2.92 (0.93) | 3.00 | 11,545.00 | − 0.976 | 0.329 | – |
02 ergonomic work-compatible furniture | 256 | 2.70 (1.08) | 3.00 | 96 | 2.97 (0.95) | 3.00 | 10,680.00 | − 1.967 | 0.049* | 0.20 |
| F2 Indoor Environmental Quality |
07 good ventilation conditions | 256 | 3.71 (0.54) | 4.00 | 96 | 3.54 (0.65) | 4.00 | 10,732.00 | − 2.311 | 0.021* | 0.20 |
04 adequate supply of daylight | 257 | 3.67 (0.62) | 4.00 | 96 | 3.59 (0.63) | 4.00 | 11,437.00 | − 1.1334 | 0.182 | – |
06 comfortable temperature conditions | 257 | 3.63 (0.58) | 4.00 | 96 | 3.52 (0.58) | 4.00 | 10,944.00 | − 1.954 | 0.051 | – |
10 attractive interior design | 257 | 3.39 (0.78) | 4.00 | 96 | 3.22 (0.76) | 3.00 | 10,671.00 | − 2.147 | 0.032* | 0.21 |
05 pleasant view | 257 | 3.26 (0.99) | 4.00 | 95 | 2.92 (1.03) | 3.00 | 9741.00 | − 3.161 | 0.002** | 0.32 |
Influence of physical home learning environment on motivation, stress and well-being
We used the WHO-5 World Health Organisation-Five Well-Being Index for measuring the overall well-being. The average well-being score measured was 14.29 (
SD = 5.58) for UWK and JKU students, where 0 is the worst possible well-being and 25 is the best possible well-being (13 and below is considered poor well-being). The average stress level measured by Perceived Stress Questionnaire (PSQ) was reported to be 40.44 (
SD = 21.85), with a range from 0 to 100, and a higher score refers to a higher stress level. Students seem to be less worried compared to the general stress level (
M = 34.06,
SD = 25.90), the level of tension was 40.33 (
SD = 25.85) and the level of demand 48.24 (
SD = 24.96). Our analysis showed that despite the lockdown, happiness was in the upper average range with 60.95 (
SD = 24.06). Motivation to learn during the COVID-19 restrictions, measured by the achievement motivation test LEIMO marker items, was reported still high (
M = 3.90,
SD = 0.78). Table
8 shows the mean values (
M) and standard deviations (
SD) for the questionnaires.
Table 8
Means and standard deviations for stress, well-being and motivation (UWK and JKU students)
WHO-5 well-being | 350 | 14.29 (5.58) |
PSQ overall | 352 | 40.44 (21.85) |
Worry | 349 | 34.06 (25.90) |
Tension | 349 | 40.33 (25.85) |
Joy | 350 | 60.95 (24.06) |
Demand | 349 | 48.24 (24.96) |
LEIMO motivation | 344 | 3.90 (0.78) |
Furthermore, we also investigated to what extent the students of UWK and the students of JKU differ in their well-being, stress perception and their motivation to learn. The difference was tested using independent-samples t-tests. Statistically significant differences were found in all scales, except for demand. When looking at the mean values, it is noticeable that UWK students have a higher sense of well-being, are more motivated to learn and have a lower stress level compared to JKU students in our sample (Table
9).
Table 9
Stress, well-being, and motivation scores by university
WHO-5 well-being | 254 | 14.99 (5.46) | 96 | 12.44 (5.51) | 3.88 (169.7), < .001** | 0.47 |
PSQ overall | 256 | 37.23 (21.56) | 96 | 48.87 (20.42) | − 4.67 (179.4), < .001** | − 0.55 |
Worry | 254 | 28.44 (22.37) | 95 | 49.09 (28.70) | − 6.33 (138.9), < .001** | − 0 .85 |
Tension | 254 | 37.78 (25.68) | 95 | 57.16 (25.19) | − 3.08 (171.6), .002** | − 0.37 |
Joy | 255 | 64.10 (24.14) | 95 | 52.49 (21.80) | 4.30 (185.2), < .001** | 0.49 |
Demand | 254 | 46.94 (25.46) | 95 | 51.72 (23.33) | − 1.66 (182.9), .099 | − .19 |
LEIMO motivation | 249 | 3.99 (.76) | 95 | 3.69 (0.82) | 3.10 (159.3), .002** | 0.39 |
In order to determine how the physical-spatial conditions influence the motivation, stress and well-being of UWK and JKU students, three hierarchical multiple regression analyses were carried out, with considering gender and institution in a first step. Mean scores of the standardised questionnaires (motivation: LEIMO; stress: PSQ; well-being: WHO-5) were calculated and used as the criteria variables. Predictors were determined as gender (male vs. female); institution (UWK vs. JKU); availability of the learning place (available all time vs. place had to be shared with others); Learning Place Quality (calculated mean score) and Indoor Environmental Quality (calculated mean score). For determining the generalisability of the regression model, we proved if the underlying assumptions have been met. The results show that none of the correlations between predictor variables exceeded the critical value of 0.80 for multicollinearity. The normal distributions of residuals were confirmed (normal curve of histogram; normal probability represents an approximately 45-degree line). The residual terms were independent since the Durbin-Watson coefficient (d) was around 2.0 (motivation: d = 1.881, stress: d = 1.954, well-being: d = 1.885).
We entered the predictors into hierarchical regression in two steps. In the first step, gender and institution were entered into regression; in the second step, availability of the learning place, Learning Place Quality and Indoor Environmental Quality were entered into the regression analyses. Table
10 shows the results for each model for each variable.
Table 10
Results of regression models of predictors of motivation, stress and well-being
Step 1 (R = .223, R2 = 5.0%, R2adj = 4.4%) |
Constant | 4.31 | 0.13 | |
Gender (female; male) | − 0.21 | 0.08 | − .014* |
Institution (UWK; JKU) | − 0.27 | 0.09 | − 0.16** |
Step 2 (R = .324, R2 = 10.5%, R2adj = 9.1%) |
Constant | 3.12 | 0.40 | |
Gender (female; male) | − 0.21 | 0.08 | − 0.14* |
Institution (UWK; JKU) | − 0.24 | 0.09 | − 0.14* |
Availability of learning place | 0.04 | 0.11 | 0.02 |
F1 Learning Place Quality (score) | 0.22 | 0.08 | 0.19** |
F2 Indoor Environmental Quality (score) | 0.13 | 0.10 | 0.08 |
Step 1 (R = .240, R2 = 5.7%, R2adj = 5.2%) |
Constant | 39.41 | 3.54 | |
Gender (female; male) | − 1.55 | 2.29 | − 0.04 |
Institution (UWK; JKU) | 11.90 | 2.60 | 0.24** |
Step 2 (R = .400, R2 = 16.0%, R2adj = 14.7%) |
Constant | 72.53 | 10.73 | |
Gender (female; male) | − 1.84 | 2.24 | − 0.04 |
Institution (UWK; JKU) | 10.55 | 2.49 | 0.22** |
Availability of learning place | 5.39 | 2.83 | 0.11 |
F1 Learning Place Quality (score) | − 4.19 | 2.20 | − 0.13 |
F2 Indoor Environmental Quality (score) | − 7.45 | 2.70 | − 0.17** |
Step 1 (R = .201, R2 = 4.0%, R2adj = 3.5%) |
Constant | 14.93 | 0.92 | |
Gender (female; male) | 0.03 | 0.59 | 0.003 |
Institution (UWK; JKU) | − 2.52 | 0.67 | − 0.20** |
Step 2 (R = .384, R2 = 14.8%, R2adj = 13.5%) |
Constant | 3.56 | 2.77 | |
Gender (female; male) | 0.15 | 0.58 | 0.01 |
Institution (UWK; JKU) | − 2.08 | 0.64 | − 0.17** |
Availability of learning place | − 0.50 | 0.73 | − 0.04 |
F1 Learning Place Quality (score) | 1.29 | 0.57 | 0.15* |
F2 Indoor Environmental Quality (score) | 2.20 | 0.70 | 0.20** |
Regression analysis for motivation shows a multiple correlation coefficient of R = 0.223 between the linear combination of the two predictors gender and institution. Model 1 statistically significantly predicts the motivation to learn, F(2,336) = 8.77, p = < 0.001, R2 = 0.050, R2adj = 0.044. The combination of these two predictors accounts of 4.4% of the variation in motivation to learn. According to standardized coefficients (β), there is a negative correlation between gender and motivation as well as between institution and motivation. This result indicates that the motivation to learn is higher for female students and UWK students. In Model 2, the multiple correlation coefficient between the linear combination of three predictors, the combination of availability of learning place, Learning Place Quality and Indoor Environmental Quality and motivation, increased to R = 0.324 after controlling for the effects of gender and institution. Model 2 statistically significantly predicted motivation, F(5,333) = 7.79, p = < 0.001, R2 = 0.105, R2adj = 0.091. The combined factors of availability of learning place, Learning Place Quality and Indoor Environmental Quality accounted for 9.1% of the variance in motivation above gender and institution. According to standardised coefficients (β), there is a still a negative relationship between gender, institution, and motivation to learn. Furthermore, Learning Place Quality showed a positive impact on motivation to learn (β = 0.19).
For stress, the multiple correlation coefficient between the linear combination of the two predictors gender and institution is R = 0.240. Model 1 statistically significantly predicted stress perception, F(2,344) = 10.49, p = < 0.001, R2 = 0.057, R2adj = 0.052. The combination of these two predictors accounts for 5.2% of the variation in stress perception. According to standardised coefficients (β), there is a positive relationship between the institution and stress (β = 0.24), which may indicate that stress perception is higher for JKU students in our sample. In Model 2, the multiple correlation coefficient between the linear combination of three predictors, i.e., the combination of the availability of the learning place, Learning Place Quality and Indoor Environmental Quality, and stress increased to R = 0.400 after controlling for the effects of the gender and institution. Model 2 statistically significantly predicted stress, F(5,341) = 12.96, p < 0.001, R2 = 0.160, R2adj = 0.147. The combined factors of availability of learning place, Learning Place Quality and Indoor Environmental Quality accounted for 14.7% of the variance in stress controlling gender and institution. According to standardised coefficients (β), there is a still a positive relationship between institution and stress (β = 0.22); and a negative relationship between Indoor Environmental Quality and stress (β = − 0.17).
Regarding the well-being, analysis shows a multiple correlation coefficient of R = 0.201 between the linear combination of gender and institution, and well-being. Model 1 statistically significantly predicted well-being, F(2,342) = 7.20, p = 0.001, R2 = 0.040, R2adj = 0.035. The combination of these two predictors accounts for 3.5% of the variation in well-being. According to standardised coefficients (β), there is a negative relationship between the institution and well-being (β = − 0.20), which indicates a lower well-being for JKU students. In Model 2, the multiple correlation coefficient between the linear combination of three predictors, i.e., the combination of the availability of the learning place, Learning Place Quality and Indoor Environmental Quality, and stress increased to R = 0.384 after controlling for the effects of the gender and institution. Model 2 statistically significantly predicts stress, F(5,339) = 11.75, p < 0.001, R2 = 0.148, R2adj = 0.135. The combined factors of availability of learning place, Learning Place Quality and Indoor Environmental Quality accounted for 13.5% of the variance in well-being above gender and institution. Standardised coefficients (β) show a negative relationship between institution and stress (β = − 0.17); and a positive relationship between Learning Place Quality (β = 0.15) and Indoor Environmental Quality and well-being (β = 0.20).
In this research, it was explored how well spatial characteristics such as the availability of the learning place, Learning Place Quality and Indoor Environmental Quality predict motivation, stress, and well-being. We also analysed the role of gender and institution. Results showed that spatial characteristics explained 9.1% in motivation, 14.7% in stress, and 13.5% in well-being after controlling for gender and institution.
In all models the institution shows an impact on motivation to learn, stress perception and well-being. Learning Place Quality shows a similar impact on motivation (β = 0.19) and on well-being (β = 0.15). The same applies to the influence of Indoor Environmental Quality, here it also shows a highly similar impact on stress (β = − 0.17) and on well-being (β = 0.20).