A total of 505 participants participated in the study. The mean age of the participants was 23.82 ± 5.74 (min: 18, max: 47). When the education level of the participants was examined, 1 (0%) participant had a primary school education, 33 (6.5%) participants had a secondary school education, 171 (33.9%) participants had a high school education, 262 (51.9%) participants had an associate’s degree, and 38 (7.5%) participants had a graduate education. When the income distribution is examined, it is found that the income of 228 (45.1%) participants is less than their expenditure, the income of 210 (41.6%) participants is equal to their expenditure, and the income of 67 (13.3%) participants is more than their expenditure. A total of 413 (81.8%) of the participants did not have any addiction, 6 (1.2%) were addicted to alcohol, and 86 (17%) were addicted to cigarettes. A total of 171 (33.9%) of the participants had a relationship, and 334 (66.1%) had no relationship. The number of participants with an active sexual life was 74 (14.7%). A total of 436 (86.3%) of the participants had knowledge about STDs. (Table 1.)
According to Table 2, the KMO value is 90.6%. A KMO test result greater than 60% and a significant Bartlett test result (p < 0.001) indicate that the scale is suitable for factor analysis. Therefore, there are correlations between the items.
In the study, the varimax rotation method was used to bring the factors together with the related items. As a result of Varimax rotation, two factors with eigenvalues greater than 1 were found in the scale: (1) Factor eigenvalue = 8.595, and (2) Factor eigenvalue = 4.146. The percentages of variance explained by both factors are 38.161 and 19.753, respectively. The total variance explained is 57.914%. Thus, it is concluded that the STD scale is two-dimensional. The factor loads of the two-factor STD behavior scale, which was created as a result of the EFA applied, and the percentage of total variance they explained are summarized in Table 3.
When Table 3 is examined, it can be said that the scale can explain attitudes towards STDs well. According to factor analysis, two subdimensions of the scale are shown in Table 3. Finally, the internal consistency coefficient of the scale consisting of 21 items was calculated. The overall Cronbach’s alpha coefficient of the scale was calculated as 0.911 for the first sub-dimension, 0.941 for the second sub-dimension, and 0.889 for the second sub-dimension. The results regarding the reliability coefficient are given in Table 4.
When Table 4 was examined, it was concluded that the internal consistency of the scale was sufficient, and the difference between the scale items was statistically significant according to the Hotelling T2 test (p < 0.001).
To evaluate the test-retest reliability of the developed scale, the scale was reapplied to 50 participants after a certain time. The consistency between the first and last test results obtained from the same 50 participants was evaluated with the intraclass correlation coefficient (ICC), and the significant differences within the group were evaluated with the Wilcoxon test. The results are given in Table 5.
When Table 5 is examined, the ICC coefficient of the test-retest results for the total score of the scale indicates that there is statistically significant and moderate compatibility. The difference between the means of test-retest results, on the other hand, does not have a statistically significant difference. As a result of the test-retest performed for the first and second subdimension scores of the scale, the ICC coefficient was found to be 1 and 0.945, respectively, and it was found to be statistically significant and highly compatible. In addition, the difference between the averages of the test-retest results of both the first and second subdimension scores is not statistically significant. In light of these results, the test-retest reliability of the STD Behavior Scale is also ensured.
Confirmatory factor analysis (CFA) was applied to examine the two-factor construct validity of the STD Behavior scale created by EFA. The model to be tested in line with examining the construct validity of the EFA results was created using 21 observed and two latent variables (Factor 1 and Factor 2). Analyses were made using the LISREL 8.7 program. The fit index values for the measurement model were found to be \({\chi }^{2}/SD\) =1.39 (p < 0.001), CFI = 0.97, RMSEA = 0.04, NNFI=0.96, NFI=0.89, SRMR=0.07. As a result of CFA, the fit index values indicate that the model and the data are compatible. The measurement model and the goodness of fit criteria for the model are given in Table 6.
