Capital Factors of Rubella Vaccination Behaviour among the Parents in Bangladesh : A PLS-SEM Analysis

This study predicts rubella vaccination behavior of the parents in Bangladesh based on CAPSES model : Human Capital (HC), Material Capital (MC) and Social Capital (SC). A sample size of 72 parents was determined to collect data. The study employed structured questionnaire technique based on 5 point-Likert scale. Then, the study used SmartPLS software (Windows Version-3) to analyze both the measurement and structural models. The study findings lead to this proposition that the higher the CAPSES ranks of the parents, the more likely they will vaccinate their children against MR. On the contrary, the lower the CAPSES rank of the parents, the less likely they will vaccinate their children against MR. Overall, the models have predictive power and relevancein the respective models (Rof RVB1= 0.862) was found to be more substantial compared to two other CAPSES models (R of RVB2 = 0.240 and R of RVB3 = 0.181). Keyword: CAPSES model, Rubella Vaccination Behaviour, Human Capital, Material Capital, Social Capital


I.1 Background of the Study
Rubella which is worldwide a major public health concern is as infectious as flu affecting all ages and sexes.Abortion is the most significant aftermath of rubella infection during pregnancy is (Gregg 1941;Miller 1994;Reef 2000;Strauss 1989).If the pregnant woman catches rubella during 1st trimester (up to week 13 of the pregnancy), there is an 80% probability that the foetus will also be infected.If she catches rubella during weeks 8-10 of the pregnancy, there is a 90% probability that the foetus will have birth defects (Azad 2012(Azad & 2014)).The rubella infection can easily pass to the foetus and cause birth defects, which is known as CRS (Congenital Rubella Syndrome), causing cataracts including other eye defects, deafness, heart abnormalities, a small head, mental retardation, a slower than normal growth rate, and the injuries in the brain, liver and lungs (Azad 2012(Azad & 2014)).In Bangladesh, rubella forms a severe problem affecting all ages and sexes.A national study shows that sera from 609 pregnant women were tested where 86% were positive in rubella IgG.The prevalence rate was 80% between 15 and 20 years of age (Nessa 2000).Nessa (2008) in their study on urban and rural Bangladeshi women showed that 71.99 percent of 582 women in the child bearing age was Rubella IgG positive.Rahman (2002) conducted a study on 198 hearingimpaired and 200 without hearing impaired children where rubella antibody was detected as positive in 74% of the hearing-impaired children and in 18% of those with normal hearing.
According to Global Alliance for Vaccination/Immunization (GAVI 2013), rubella incidence in Bangladesh was found among 37 per million individuals followed by 87 per million.WHO (2012) predicts that MR vaccination campaign can significantly reduce the rubella incidence.
According to a report of WHO (2009), all WHO regions either controlled or eliminated MR.In August 2012, the Expanded Programme on Immunization in Bangladesh enlisted rubella vaccine combined with measles to get protected against the disease.Like many countries in Asia, the uptake of MR vaccine was relatively low (WHO 2012;DGHS 2013).The Ministry of Health and Family Welfare of the Government of Bangladesh, to increase the coverage, implemented measles-rubella (MR) vaccination campaign in 2014 among the children aged 9 months to 15 years.During the campaign directed by the GAVI, 86 percent MR vaccination coverage was found among the children of 9 months to 15 years of age (Uddin 2016).However, Azad (2014) found merely 1 % students (whose age exceeded 18 years) of Northern University, Bangladesh to be vaccinated against MMR.
To understand why MR vaccination coverage varied in terms of age, the obvious reason may be the restriction of the free vaccine only up to age 15 years.So, now is the time to conduct research on MR vaccination coverage on the adults of Bangladesh depending on three kinds of capital: material capital (MC), human capital (HC) and social capital (SC) what Oakes & Rossi (2003) called together CAPSES scale.However, no study has been conducted on explaining rubella vaccination behaviour in Bangladesh based on the CAPSES scale.So, this is the high time to examine the usefulness of CAPSES model to understand the phenomena of rubella vaccination behaviour.

I.2 Research Goal and Objectives
The main goal of the research is to predict rubella vaccination behaviour among the parents in Bangladesh based on three capital factors of socio-economic status (CAPSES Model).In this regard, the study attempts:  to understand rubella vaccination behaviour among the study parents;  to measure the capital factors of the parents in the study areas; and finally  to examine the effects of capital factors on rubella vaccination behaviour of the study parents.

CAPSES Model of Health Behaviour
No studies have hitherto been conducted on explaining rubella vaccination behaviour among the parents of Bangladesh using CAPSES model.However, many studies have been conducted on explaining the health behaviour based on CAPSES model in global context.Grossman and Kaestner (1997) found the effect of education as human capital on health status.Parental educational status was also found to be powerful predictor of child health (Desai and Alva 1998;Frost 2005;Thomas 1990Thomas , 1991;;Mosley and Chen 1984;Chou 2007).Fewer studies reported positive associations between maternal schooling and child health (Shakotko 1981).Oakes and Rossi (2003) suggested material capital to understand health behaviour.(Robert & House 1996;Spilerman 2000).Many systematic literature reviews (42 papers) showed that health behaviour was significantly associated with social capital (Islam 2006).Kim's (2008) study findings revealed strong relation of physical health with social capital.Wen (2005) conducted a survey on patients diagnosed in 1993 and found that mortality was associated with contextual collective efficacy.Recently, Oakes & Rossi (2003) recommended CAPSES scale for conducting social scientific behavioural researches.
Though there are considerable works on the health issues based on CAPSES model in the global level, there is a gap of research on conducting rubella vaccination behaviour using CAPSES model in Bangladeshi context.

1.4Conceptual Framework
Based on the research objectives and questions as well as considering the gap of the existing literature, the study has designed the following proposed model:

Rubella Vaccination Behaviour
Rubella vaccination behaviour will be measured by offering 12 statements where the respondents will be asked to choose an option under 5 point-Likert scale.

Human Capital (HC)
The study defines human capital as individual competency, knowledge, physical abilities (Oakes and Rossi 2003) and academic experience/achievement (Lyons 2003).

Material Capital (MC)
In this study, material capital has been defined as the real property (e g. home land), income from earning e g. wages/salaries, tips and other taxable employee pay/selfemployment, income from investment e g. interest payments, dividends and capital gains collected upon the sale of a security and fungible goods e g. gold, company shares, bonds, other precious metals and currencies (Oakes and Rossi 2003).

II.
Research Methodology

Questionnaire Design
For collecting data, the study conducted a survey consisting of 4 sections.The first section will cover the respondents' socio-demographics such as gender, age, family size, religion and ethnicity.The other sections of the questionnaire will be: (1) Social Capital (SC); (2) Material Capital (MC); (3) Human Capital (HC); and (4) Rubella Vaccination Behaviour in the study areas.For quantification and scorekeeping, the study will use a 5 point-Likert Scale (e.g., Strongly Disagree=1, Disagree=2, Neutral=3, Disagree=4 and Strongly Agree=5) for better understanding of the magnitude of the respondents' behaviour.

Participants
The study which is quantitative in nature was carried out on 72 parents of Bangladesh.The sample size was determined considering 12 indicators (6 samples per indicator).Among three categories of the respondents: high awareness, low awareness and no awareness of rubella vaccination, 24 respondents from each category were confirmed for data analysis.However, the respondents were taken from 2 districts-Rajbari (one of the highest vaccine coverage areas) and Munshiganj (one of the lowest vaccine coverage areas)-each containing 36.

II.3 Research Instrument and Statistical Analysis
After collecting data from the field, to analyse both measurement model and the structural model, the study used Smart PLS software.To evaluate former model, average variance extracted (AVE), Fornell-Larcker criterion and Heterotrait-Monotrait (HTMT) ratio were evaluated (Wong 2013;Hair et al. 2010Hair et al. , 2011Hair et al. and 2014)).Then, the structural model was evaluated through collinearity assessment, standardized path coefficients, T-statistics/P values, coefficient of determination (R 2 ), effect size of exogenous constructs on endogenous constructs (f 2 ), predictive power/relevance of endogenous construct (Q 2 ) and predictive contribution of exogenous to endogenous constructs (q 2 ).

Ethical Consideration
Voluntary sharing of the respondents as well as confidentiality of their information were strictly maintained.While interviewing the respondents, force and coercion was avoided and their privacy was safeguarded.

III.
Results of the Study

Characteristics of the Respondents
Table 1 shows that 50 percent of the respondents were above 16 years and the remaining were upto 16 years.In terms of ethnicity and religion, 100 percent respondents were found Bengali and Muslim. Budapest International Research and Critics Institute-Journal (BIRCI-Journal) Volume I, No 3, October 2018, Page: 56-70 e-ISSN: 2615-3076(Online), p-ISSN: 2615-1715(Print) www.birci-journal.com emails; birci.journal@gmail.com birci.journal.org@gmail.com

Respondents' Status of Rubella Vaccination Behaviour
The statuses of the respondents in terms of rubella vaccination behaviour are evident from the Table 2 below:

Results of Measurement Models
The combined model (Figure 2) shows that there is a positive effect of Hum Cap, Soc Cap and Mat Cap on RVB1 while the negative effects of those constructs on RVB2 and RVB3 were reported.To compare those models, it is necessary to develop separate models (Model RVB1, Model RVB2 and Model RVB3).
First, 4&5 show the internal consistency and reliability of data where the values are higher than the threshold value of 0.70 (Wong 2013; J. F. Hair et al. 2010Hair et al. , 2011Hair et al. , 2014Hair et al. , 2016;;Nunnally and Bernstein 1994;Urbach and Ahlemann 2010).Convergent validity is also confirmed since the average variance extracted (AVE) of each latent variable is found to be much greater than the acceptable threshold of 0.5 (Wong 2013; J. F. Hair et al. 2010Hair et al. , 2011Hair et al. and 2014;;Nunnally and Bernstein 1994;Urbach and Ahlemann 2010 (Table-4).

Results of Three Comparative Structural Models
To evaluate structural model, the study considers collinearity assessment, path coefficients, T-statistics and P values, coefficients of determination (R 2 ), effect size of exogenous constructs on endogenous constructs (f 2 ), predictive power/relevance of endogenous construct (Q 2 ) and predictive contribution of exogenous to endogenous constructs (q 2 ).First, in terms of collinearity, no VIF values are found to be greater than 3 .So, the latent variables have no problem of multi-collinearity (J.F. Hair et al. 2010Hair et al. , 2011Hair et al. and 2014;;Kock and Lynn 2012;Diamantopoulos and Siguaw 2006).
Thirdly, in terms of path coefficient effect, we find from figure-4a, 4b & 4c an interesting homogeneous report that Soc Cap has the strongest effect on rubella vaccination behaviour followed by Hum Cap and Mat Cap among the three categories of respondents.So, the hypothesized path relationships between the predictors and Rubella Vaccination Behaviour are found to be statistically significant since the standardized path coefficients are greater than the threshold value of 0.1 (Wong 2013).However, only the CAPSES model of RVB1 show that the three constructs have positive effect on Rubella Vaccination Behaviour.On the contrary, the remaining two CAPSES models show the negative effect of those constructs on Rubella Vaccination Behaviour.
Fourthly, in terms of structural path significance based on bootstrap method taken the large number of subsamples of 5000, T-statistics and P values using a two-tailed t-test with a significance level of 5%, we find that except the linkage between Soc Cap and RVB2 (T value of 1.933 & P value 0.053), all the linkages are statistically significant since all the T-statistics are greater than 1.96 and the P values are smaller than 0.05 (Table -7).
Fifthly, in terms of coefficient of determination, R 2 , the endogenous latent variable substantially explain 86.2 % of the variance in RVB1 whereas in cases of two other models, R 2 values are not strong (R 2 for RVB2= 0.240 & for RVB3=0.181)(see Wynne W. Chin 1998Chin , 2010)).Sixthly, the study finds the effects (f 2 ) of Hum Cap, Mat Cap and Soc Cap on the 'RVB1' are very strong but weak on 'RVB2' & 'RVB3' (endogenous latent variable) .
Seventhly, in addition to evaluate the magnitude of R 2 values as criterion of predictive accuracy, the study has also examined Stone-Geisser's Q 2 value (Geisser 1975;Stone 1974).That is, to measure the model's predictive power or predictive relevance of endogenous construct 'Rubella Vaccination Behaviour', the value of Q 2 has been calculated using Blindfolding method.The study finds the strong predictive power and relevance of endogenous constructs 'RVB1' since the Q 2 value is 0.841 that is much greater than the threshold value of 0 (Geisser 1975;Stone 1974).However, the predictive power and relevance of endogenous constructs 'RVB2' and 'RVB3' were found to be comparatively much weaker since the values of Q 2 were respectively 0.217 and 0.161 .

IV. Conclusion and Recommendations
The study explains and predicts Rubella Vaccination Behaviour among the three categories of parents in Bangladesh based on the CAPSES model: highly aware, low aware and not aware of rubella vaccination.After PLS-SEM analysis of empirical data, the study finds significant effects of human capital, material capital and social capital on the Rubella