Mean age of the participants was 39.24±11.73 and mean history of using drugs was 13.8±11.04. Totally, 86.6% were men, 54.8% were married, 43.6% had an elementary level education, 40.4% had a high school diploma. Moreover, 29.2% had used opiate, 22.4% used heroin and crack heroin, and 48.4% used a combination of natural and industrial opiate. Additionally, 91.6% lived in urban area.
Validation of the tool
The first step to check validity of the tool is content validity check. Waltz and Bausell’s CVI was employed to this end. As the results showed, the CVI and CVR were acceptable for all the statements and no statement was omitted at this stage. To examine reliability of the tool, test-retest technique was used through Pearson’s correlation coefficient, which yielded 0.875.
To examine construct validity, exploratory factor analysis was used followed by confirmatory factor analysis. In the former, correlation coefficients were examined for the statements to make sure that they are in an acceptable range. Kaiser Meyer Olkin (KMO) test and Bartlett’s test of sphericity were used to this end. Given that KMO = 0.858 >0.7 and that Bartlett’s test was significant (Chi Square = 13500/19, P-value <0.01), the presumptions for using exploratory factor analysis on TCU questionnaire with 83 statements were met. Varimax vertical rotation was employed and the factors of which the specific value was above one were selected for exploratory factor analysis through principle components (PC) analysis. In this study, factors with eigenvalues greater than one were selected (Figure 1).
In addition, commonality value of each statement was high (>0.5) so that none of the questions were omitted in this stage. Still, factor load of the rotated variables showed that some of the variables had factor loading (>0.3) on two factors at the same time and therefore, they were omitted. In this way, 24 statements (1, 2, 10, 12, 14, 15, 19, 23, 27, 28, 32, 35, 37, 47, 48, 51, 54, 56, 59, 68, 70, 74, 75, and 81) were omitted. In addition, statement No.26 was omitted because of low factor loading (<0.3) on different factors. Thus, 57 statements remained in the study. Exploratory factor analysis was repeated using the main elements of the analysis and varimax rotation. Scree plot demonstrates factor analysis in SPSS so that 13 factors or elements are fitted for the final analysis (Table 1). The questions about each factor, name of each factor, and Cronbach’s alpha coefficients are listed in Table 2 to determine reliability of the elements. Exploratory factor analysis was completed with 11 factors and 56 statements.
First-order confirmatory factor analysis was used in this study in two steps. In the first step, factor loadings of the questionnaire questions were analyzed. Secondly, factor loadings of factors were analyzed (Table 3). Only the statement No. 46 had a low factor load (t=0.26) and eliminated.
Analysis of the relationships
As listed in Table 4, the variables gender, job, income, and marital status had a relationship with the psychological functioning of the PWUDs (P<0.05). However, education level, domicile, the way of using drugs, and the type of drugs did not have a significant relationship with one’s social functioning (P>0.05).
Table 5 lists the Pearson correlation coefficients to examine the relationship between demographical variables, psychological functioning, and its aspects.
As listed in Table 5, there is a significant relationship between age, number children, and the history of using drugs and psychological functioning (P<0.01). In other words, with an increase in the demographical variable, a decrease in psychological functioning takes place. However, there was no relationship between the age of first experience of drugs and psychological functioning (p=0.513). Table 6 compares mean score and SD of the aspects of social functioning in terms of demographical variables.
As the findings showed, domicile, job, income, and marital status had a relationship with social functioning of the patients (P<0.05). However, gender (p=0.674), education level (p=0.432), way of using drug (p=0.431), and type of drug (p=0.739) did not have a significant relationship with social functioning. Table 7 lists Pearson correlation coefficients for the relationship between demographical variables, psychological functioning, and its aspects.
As listed in Table 7, there is a negative significant relationship between age and violence in the patients (p<0.01). In other words, with an increase in age, violence declines in the patients. There was no significant relationship between other demographical variables and social functioning (p>0.05). As listed in Table 8, none of the demographical variables are related to the motivation for treatment in the subjects (p>0.5). Only marital status was significantly related to treatment readiness. So that, the widowers/widows had more motivation to quit. In addition, the type of drug has a significant relationship with treatment readiness (P<0.05); so that patients who use only one type of drug have more desire for treatment.
As listed in Table 9, there is a negative relationship between number of children and motivation for treatment (p<0.05, r=-0.139). That is, with an increase in the number of children, the motivation in patients declines.