4) Highest Education Level No formal education 1(4) Primary school 10(40) Secondary school 8(32) Form 6/A Level (or equivalent) 3(12) Degree 2(8) Postgraduate 1(4) Work Private sector 1(4) Government sector 0(0) Own business 1(4) Housewife 15(60) Student 0(0) Retired 7(28) Others 1(4) Monthly FT011 manufacturer Salary No income/not relevant 20(80) MYR 1- MYR 2000 4(16) MYR 2001- MYR 4000 0(0) MYR 4001- MYR 6000 1(4) > MYR 6000 0(0) Number of medicine items 1-3 items 23(92) 4-6 items 2(8) 7-9 items 0(0) MYR = Malaysian ringgitkappa statistic for nominal scale and Intraclass Correlation Coefficient (ICC) was applied in interval scale. The ICC model used is the One-way random 31 effects model, single measures. Confirmatory factor analysis (CFA) was performed to assess the construct validity of PVASQ. CFA examined the construct validity using principal axis factoring and Varimax rotation with Kaiser normalization. The TPB postulates three conceptually independent determinants of 18 Therefore, the decision to choose intention. 32 Varimax rotation is well supported. Instead of relying on eigenvalues to provide the domains for our constructs, we restricted the extracted factors to 33 a fixed number. In this case, we restricted the extracted factors to four based on the three TPB constructs of attitudes, subjective norms and perceived behavioral control and the fourth construct of expectations. This is therefore consistent with our TPB model and the findings of our initial qualitative investigation which revealed the role of expectations. Items in the intention scale were not included into factor analysis because intention is a dependent variable in the TPB model. The knowledge scale is a dichotomous variable which is unsuitable for factor analysis simply because binary variables cannot be expressed 34 Thus, the Intraclass within factor models. Correlation Coefficient by test-retest is sufficient to indicate construct validity for the knowledge scale. RESULTS Reliability and ICG-001MedChemExpress ICG-001 Internal Consistency A total of 25 respondents participated in the study, and there were more women (80 ) than men (20 ) in the sample. A majority of the respondents were Chinese and aged from 41-50 (36 ), 51-60 (36 ) and older than 60 (24 ). About 40 of respondents completed primary school, 32 had secondary schooling, and 12 completed a high school education at the equivalent of A-Levels. Sixty percent of the participants were housewives and 28 were retired citizens. There were twenty respondents (80 ) reported not having any income while 16 had a monthly salary between MYR 1MYR 2000. Respondents who received pocket money from their spouses were included in the “no income or not relevant” category. Respondent’s demographic and characteristics are shown in Table 1.Stage 6: Statistical Analysis: Internal Consistency, Test-retest reliability and construct validity Data entry and analysis was performed using SPSS version 22. The significance level was set at p<0.05. Demographic data was presented in numbers and percentages. Internal consistency was assessed using Cronbach's Alpha. The stability of the construct measures were assessed by test30 retest reliability. The chance correlated agreement reliability for twenty five respondents at two time points was calculated using Cohen's unweightedTable 2. Cronbach's Alpha values for all scales and TPB constructs. Cronbach Alpha Coefficient Number of items in Domain Test (Week 1) Retest (Week 2) each Scale N = 25 N = 25 Knowledge scale 7 0.763 0.7.4) Highest Education Level No formal education 1(4) Primary school 10(40) Secondary school 8(32) Form 6/A Level (or equivalent) 3(12) Degree 2(8) Postgraduate 1(4) Work Private sector 1(4) Government sector 0(0) Own business 1(4) Housewife 15(60) Student 0(0) Retired 7(28) Others 1(4) Monthly Salary No income/not relevant 20(80) MYR 1- MYR 2000 4(16) MYR 2001- MYR 4000 0(0) MYR 4001- MYR 6000 1(4) > MYR 6000 0(0) Number of medicine items 1-3 items 23(92) 4-6 items 2(8) 7-9 items 0(0) MYR = Malaysian ringgitkappa statistic for nominal scale and Intraclass Correlation Coefficient (ICC) was applied in interval scale. The ICC model used is the One-way random 31 effects model, single measures. Confirmatory factor analysis (CFA) was performed to assess the construct validity of PVASQ. CFA examined the construct validity using principal axis factoring and Varimax rotation with Kaiser normalization. The TPB postulates three conceptually independent determinants of 18 Therefore, the decision to choose intention. 32 Varimax rotation is well supported. Instead of relying on eigenvalues to provide the domains for our constructs, we restricted the extracted factors to 33 a fixed number. In this case, we restricted the extracted factors to four based on the three TPB constructs of attitudes, subjective norms and perceived behavioral control and the fourth construct of expectations. This is therefore consistent with our TPB model and the findings of our initial qualitative investigation which revealed the role of expectations. Items in the intention scale were not included into factor analysis because intention is a dependent variable in the TPB model. The knowledge scale is a dichotomous variable which is unsuitable for factor analysis simply because binary variables cannot be expressed 34 Thus, the Intraclass within factor models. Correlation Coefficient by test-retest is sufficient to indicate construct validity for the knowledge scale. RESULTS Reliability and Internal Consistency A total of 25 respondents participated in the study, and there were more women (80 ) than men (20 ) in the sample. A majority of the respondents were Chinese and aged from 41-50 (36 ), 51-60 (36 ) and older than 60 (24 ). About 40 of respondents completed primary school, 32 had secondary schooling, and 12 completed a high school education at the equivalent of A-Levels. Sixty percent of the participants were housewives and 28 were retired citizens. There were twenty respondents (80 ) reported not having any income while 16 had a monthly salary between MYR 1MYR 2000. Respondents who received pocket money from their spouses were included in the “no income or not relevant” category. Respondent’s demographic and characteristics are shown in Table 1.Stage 6: Statistical Analysis: Internal Consistency, Test-retest reliability and construct validity Data entry and analysis was performed using SPSS version 22. The significance level was set at p<0.05. Demographic data was presented in numbers and percentages. Internal consistency was assessed using Cronbach's Alpha. The stability of the construct measures were assessed by test30 retest reliability. The chance correlated agreement reliability for twenty five respondents at two time points was calculated using Cohen's unweightedTable 2. Cronbach's Alpha values for all scales and TPB constructs. Cronbach Alpha Coefficient Number of items in Domain Test (Week 1) Retest (Week 2) each Scale N = 25 N = 25 Knowledge scale 7 0.763 0.7.