3.4. Study Results
The survey sample primarily consisted of individuals aged 60 and older, comprising 38.1% of the total respondents. The gender distribution indicated that slightly less than half (approximately 53%) of the respondents were women (
Table 1), with the majority (87.2%) of them possessing at least a secondary level of education. Conversely, about 76% of the male respondents held a similar educational qualification. Regarding employment status, the largest segment of respondents was actively engaged in economic activities (44%), followed by those in retirement (28.2%). Among individuals aged 30 to 59, households typically comprised four members (62.4%), whereas among older individuals aged 60 and above, two-person households were predominant (45.2%).
Contingency tables were constructed to analyse the relationships between leisure use variables and socio-demographic variables (
Appendix A,
Table A1,
Table A2,
Table A3,
Table A4,
Table A5,
Table A6,
Table A7,
Table A8,
Table A9,
Table A10,
Table A11,
Table A12,
Table A13,
Table A14,
Table A15 and
Table A16). The strength of the relationships was determined using Cramer’s V coefficient, with the coefficient values provided in the table titles. Initially, we explored the connections between gender and leisure time utilisation variables (
Table A1,
Table A2,
Table A3,
Table A4,
Table A5 and
Table A6). The findings revealed a pronounced association between gender and housework, while moderate relationships were observed with walking, crossword puzzles, computer games, and meetings/chatting with friends/family via the Internet/phone. Notably, the first three activities were predominantly favoured by women, while the latter two were more popular among men.
Age emerged as a significant determinant influencing the leisure activities of adults (
Table A6,
Table A7,
Table A8,
Table A9,
Table A10,
Table A11,
Table A12 and
Table A13). Activities such as walking, working in the allotment, reading books, watching TV, and housework held greater appeal for older individuals (aged 60+). These same activities also garnered interest among respondents aged 30–44, who also preferred cycling. Conversely, running was primarily an activity that respondents aged 30–44 embraced.
Moreover, a preference for reading books and watching TV was observed among respondents with at least a secondary level of education (
Table A13,
Table A14 and
Table A15). Furthermore, those with such an educational background tended to spend their leisure time in neighbouring/German municipalities (
Table A16).
Correspondence analysis was used to search for associations between variables characterising respondents’ leisure time by constructing Burt’s matrix, 78 × 78 in dimension. This matrix was derived from thirty-six variables, each categorised in alignment with the characteristics of the research material. The dimension of the actual co-occurrence space of responses to the analysed questions was determined to be 42, based on Formula (2).
Subsequently, we evaluated how eigenvalues in a lower-dimensional space accounted for the total inertia (λ = 11,667) (The total inertia is the sum of K eigenvalues, where K is the dimension of the real co-occurrence space). To ascertain their relevance for the study, we applied Greenacre’s criterion, which considers principal inertias greater than
to be significant.
Table 2 illustrates that these relevant inertias (
Table 2 omits the results for K > 15, as for these dimensions the main inertias were no higher than 0.0278, so these dimensions were not significant in the study) are associated with K values up to 15. To enhance the quality of the mapping, we performed a modification of the eigenvalues following Greenacre’s proposal (Formula (3)).
Table 2 presents the eigenvalues
(squares of the singular values
), the contribution of the principal inertia to the total inertia (percentage inertia
) and the contribution of the eigenvalues from the K dimension to the total inertia (cumulative percentage
) before and after modification. In addition, a plot of the eigenvalues was drawn (
Figure 2). The trend observed in the curve connecting the eigenvalues suggests that the presentation space depicting the co-occurrence of variable categories should not exceed six dimensions. These six dimensions collectively capture over 60% of the inertia associated with the analysed data array.
The new coordinate values in the six-dimensional space for the variable categories were determined using the formula:
where:
—the matrix of new coordinate values for categories of variables (dimension 42 × 6),
F*—the matrix of original coordinate values for categories of variables (dimension 42 × 6),
—the diagonal inverse matrix of singular values (dimension 6 × 6), —the diagonal matrix of modified eigenvalues (dimension 6 × 6).
Interpreting results within a six-dimensional space can be challenging. Hence, we employed Ward’s method [
47,
48,
49,
50], which facilitated identifying connections between variable variants (
Figure 3). The critical distance value at which the merging of clusters ceased was determined using the metric proposed by T. Grabiński [
51]:
—the length of the i-th bond (i-th branch of the tree).
Figure 3.
Diagram of hierarchical classification of categories of variables made using Ward’s method (all respondents). Source: own elaboration.
Figure 3.
Diagram of hierarchical classification of categories of variables made using Ward’s method (all respondents). Source: own elaboration.
Based on the derived clusters, we can elucidate associations among the categories of the analysed variables, thereby unveiling patterns in respondents’ leisure time utilisation. These patterns can be summarised as follows (the categories of variables given in the
3.2 Statistical Material section are given in brackets):
Group I (M, AG1, AG2, ED1, ED4, P1, P4, P5, P2.3:1, P2.4:1, P2.6:1, P2.7:1, P2.8:1, P2.1:10, P2.1:11, P3.1:1, P3.3:1, P3.4:1, P3.7:1, P3.8:1, P3.10:1, P3.12:1, G2) primarily comprises men aged up to 44 years, with a mix of primary and tertiary education attainment (16%). They predominantly reside in single-person households and households with four or more occupants. This cluster exhibits a penchant for active leisure time pursuits, including activities such as running, swimming, engaging in physical exercise at home, cycling, participating in indoor and outdoor team games, as well as utilising leisure and sports facilities. Additionally, they are inclined to read books, indulge in board and computer games, and devote their free time to interacting with their children through play and educational support. Social gatherings with family and friends appeal to this group, as does browsing the Internet and engaging in social networks. Notably, individuals in this cluster tend to eschew spending their leisure time in neighbouring or German municipalities.
Group II (K, AG3, AG4, ED3, ED4, P2, P2.1:1, P2.2:1, P2.12:1, P3.2:1, P3.5:1, P3.11:1, P1.5:3, P1.5:4, P1.5:5, P1.7:3, P1.7:4, P1.7:5, P1.8:3, P1.8:4, P1.8:5, P1.10:3, P1.10:4, P1.10:5, G1) comprises women aged 45 and over with vocational and secondary education (32%), most often living with another person (spouse, partner). They prefer leisure activities such as Nordic walking, engaging in the allotment and manual work, watching TV, and solving crossword puzzles. Notably, they do not express a strong interest in developing green areas within their own municipality, the overall safety of their municipality, the cleanliness of recreational spaces, or the opportunity to socialise with others. Interestingly, individuals in this cluster are inclined to spend their leisure time in neighbouring and German municipalities.
Group III (variables from P1:1, P1:2 to P10:1, P10:2) pertains to individuals for whom the elements influencing leisure activities within the municipality hold significance (exceeding 50%), yet they do not prefer any leisure activity.
Employing correspondence analysis to uncover associations between variables enabled the identification of three distinct groups of respondents characterised by their leisure time utilisation. Considering the study’s objectives, the focus was on respondents aged 60 and above. From the initial sample of 1068 participants, 407 individuals within this age bracket were selected. Subsequent application of correspondence analysis in conjunction with Ward’s method (
Figure 4) facilitated the identification of patterns related to the leisure time habits of seniors. The following three groups of respondents emerged (the categories of variables given in the
3.2 Statistical Material section are given in brackets):
Group I (ED2, ED3, ED4, P2.2:1, P2.12:1, P2.8:1, P3.1:1, P3.9:1, P3.10:1, P1.5:3, P1.5:4, P1.5:5, P1.7:3, P1.7:4, P1.7:5, P1.8:3, P1.8:4, P1.8:5, P1.10:3, P1.10:4, P1.10:5, G1) encompasses respondents with a minimum level of vocational education. In their leisure time, they undertake the following activities: reading books (39%), working on the allotment (32.2%), cycling (27%), Nordic walking (9.3%), and housework and meeting with friends and family. As for their preferences, they do not attach significant importance to factors such as the development of green areas within the municipality, the overall safety of the municipality, the cleanliness of recreational spaces, or the opportunity to interact with others. Instead, they prefer to spend their leisure time in neighbouring or German municipalities.
Group II comprises individuals with a primary education background (2.2%) who do not specify any particular leisure activities.
Group III (P2.1:1, P2.1:10, P2.3:1, P2.5:1, P2.6:1, P2.7:1, P3.3:1, P3.4:1, P3.6:1, P3.7:1, P3.8:1) includes individuals with varying levels of education who express a keen interest in a diverse range of activities, with a predominant focus on physical pursuits. These activities include walking (60%), engaging in play with children (10%), running (9%), assisting children with their studies (6%), and swimming (5%). In contrast, other activities like exercising in leisure and sports facilities, working out at home, participating in indoor team games, board games, sunbathing, and computer games are of less interest, each garnering participation from less than 5% of older respondents. Notably, respondents in this group attribute significance to elements (variables from P1:1, P1:2 to P10:1, P10:2) that influence leisure time management within the municipality.