The Moderating Effect of Socio-Demographic Characteristics on Subjective Wellbeing

The independent association between socio-demographic characteristics and subjective well-being is well studied, however, the moderating impact of those characteristics are yet to be established in literature. This study examines the moderating influence of socio-demographic characteristics on subjective well-being among residents in Ibadan metropolis. Using a descriptive survey design and a multistage sampling technique, 10 enumeration areas each were selected from the five major Local Government Areas (LGAs) with simple random technique. The number of participants on the selected enumeration areas were determined with enumeration area maps. Two hundred and twenty participants each were selected from each LGA, using a systematic technique, making a total of 1,100 participants. A questionnaire focusing on socio-demographic profile and life satisfaction scale (r=0.74) was administered to the participants. Data were analyzed using descriptive statistics, co-relational analysis, moderated hierarchical multiple regression and analysis of variance at 0.05 level of significance. Two hypotheses were tested. The result reveals that socio-demographic factors jointly predicted life satisfaction (F (1, 1007) =4.61, p<.001). Age and marital status also interacted to predict life satisfaction (β=-0.13, t=-3.58, p<.01; β = -0.08, t=1.98, p<.05). The interaction effect of social demographic factors should be considered in a comprehensive examination of subjective wellbeing.


Introduction
Subjective well-being, also known as life satisfaction or happiness or quality of life International Journal of Social Science Research ISSN 2327-5510 2017 (Veenhoven, 2009) has been widely studied in psychology and other disciplines. However, studies of subjective wellbeing from different disciplines have recorded different findings across nations (Veenhoven, 1993). This suggests that the association between subjective well-being and its determinants varies from place to place. In an attempt to predict subjective wellbeing, some recent studies in Nigeria have mainly focused on psychological factors (Oladipo & Olapegba, 2012), emotional intelligence and social support (Kolade & Dennis, 2015), personality factors (Mayungbo,2016) etc. These studies have sometimes reported different and conflicting results. This is an indication that the determinants of subjective wellbeing are not universal but vary with culture, time and place (Nordbakke & Schwanen 2013).Only few studies have considered the influence of socio-demographic factors on subjective wellbeing among Nigerian population and even the few foreign studies that have done so, have focused on only one or two factors at a time i.e age and subjective well-being (Ulloa, Møller, & Sousa-Poza, 2013), gender and happiness (Zweig, 2015) and so on. There are only few studies that have directly examined the moderating effect of multiple socio demographic characteristics on subjective wellbeing among Nigerian samples. As observed by Agbo, Nzeadibe and Ajero (2012), there are limited studies on the correlates of life satisfaction of Nigerians. The purpose of this paper is to examine the moderating influence of age, gender, marital status and educational qualification on subjective wellbeing.
Research on the study of the relationship between subjective wellbeing and age have produced conflicting results. While some findings have found no relationship, other studies have reported a positive association. Old age is usually associated with reduction in human health and quality of life (Bond & Corner, 2004). Yet, some findings have indicated that subjective well-being is either stable or increasing as individuals attain old age (Blanchflower & Oswald, 2008;Frijters & Beatton, 2012).Cross-sectional studies from various cultures also conclude that life satisfaction is relatively stable across age groups in most countries (Schilling, 2006;Baird, Lucas, & Donnellan, 2010;Hansen & Slagsvold, 2011). Three suggestions were offered for individuals reporting high subjective wellbeing in old age. The first is the influence of personality and adaptation (Diener & Suh, 1998). It is assumed that even if life circumstances alter subjective wellbeing for a while, individuals would return to their baseline level over time, based on their personality traits. The second explanation suggests that the elderly seems to reduce their needs, aspirations and standard of comparison (George, 2006) and the reduction is believed to promote well-being by reducing the aspirations and achievement gaps among the elderly than the middle-aged adults (Cheng, 2004). The third explanation relates to the experiences and competencies of the elderly to regulate their emotions.
In the views of Carstensen (1991) human social goals change with age. As individuals advance in age, their desire to relate with friends and acquaintances and the satisfaction derived from such relationships decrease. Therefore old people selectively reduce their social networks and focus their time on emotionally intimate social relationships, such as close friends and family. Bowling (2011) asserts that compared to people below the age of 65, individuals above the age of 65 report having a smaller number of individuals to turn to for physical and emotional support during difficult times. However, it is not clear whether the International Journal of Social Science Research ISSN 2327-5510 2017 smaller number of individuals available in the social network is due to old people's choice or to their circumstances. Whatever the reason, the decline in social relationship and social goals suggest there may be age differences in how people report their social support experiences and their life satisfaction. It is clear that social relationship is an important aspect of life satisfaction in old age.
Gender differences in subjective wellbeing are not generally found to be in the support of women. There are gender differences in many dimensions of wellbeing across the world (Klasen, 2004) which in most cases, does not favour women. Considering the time spent in unpaid care work, women seem to have longer working weeks than men (Eurofound, 2013).Women report higher score of depression than men (Tesch-Romer, Motel-Klingebiel, & Tomasik, 2008) and they also rate their subjective health lower than men (UN World Happiness Report, 2012). Boarini, Comola, Smith, Manchin and de Keulenaer (2012) also reported the higher happiness and lower affect balance of women. Therefore, Women generally live in worse conditions than men, but Women are more satisfied with life when they are placed in the same conditions with men (Easterlin, 2003;Alesina, DiTella, & MacCulloch, 2004;Kahneman & Kruger, 2006;Kapteyn, Smith, & Van Soest, 2009;Boarini et al., 2012). Graham and Chattopadyay (2013) also found that women have higher levels of wellbeing than men, with a few exceptions in low income countries, such as Sub-Saharan Africa. On the contrary, however, Moksnes and Espnes (2013) found a lower level of life satisfaction for women than men. Similarly, Goldbeck, Schmitz, Besier, Herschbach and Henrich (2007) document a lower life satisfaction for girls as compared to boys.
Research evidence has also indicated that marriage improves the level of life satisfaction, across countries. The importance of marriage has been attributed to its provision of social support (House, Landis, & Umberson, 1988). Marriage is a social contract that binds people together in an intimate relationship that can help them cope in times of stress, marriage foster social integration and a sense of belonging and purpose to the married (Waite and Gallagher 2000). Family members are expected to provide emotionally support and are likely to contribute to well-being by providing relieve during stressful life events (Umberson & Chen, 1994). Frey and Stutzer (2002) contend that marriage provides additional sources of self esteem, social support and companionship. Lucas (2003) suggests that individuals experience changes in life satisfaction in the year after marriage. Some people report rapid reduction in wellbeing after marriage, others return to their baseline after some years while others continue to experience increase in life satisfaction over time. Yet, almost everyone experience an increase in life satisfaction immediately after marriage. Easterlin (2003) also notes increases in life satisfaction among those who married in the first decade of adulthood. However, numerous studies have suggested that married people are more satisfied with life than the never-married or previously married people (Heliwell, 2003;Strine, Chapman, Balluz, Moriarty, & Mokdad, 2008;Mousavi, Shiani, Mohammadi, Sadjadi, Tabatabaee, & Assari, 2011;Sun, Chen, Johannesson, Kind, & Burström, 2016).
Similarly , Verbakel, (2012) notes that married people are happier and more satisfied with their lives than the unmarried. The positive association between marital status and life satisfaction has also been observed by Diener, Suh, Lucas, & Smith (1999), in several cross-sectional International Journal of Social Science Research ISSN 2327-5510 2017 studies. These findings are in line with the generally held believe that, marriage can directly enhance personal well-being by providing emotional and financial support. According to Carr and Springer (2010), the influence of marriage for physical and emotional wellbeing of individuals are widely reported. However, recent studies have identified that the influence of marriage is dependent on the quality of the marriage. Marriages that are problematic are likely to have emotional effects on married individuals, while a high-quality marriage tend to provide benefits, especially for women (Proulx, Helms, & Buehler, 2007) and older adults (Umberson, Williams, Powers, Liu, & Needham, 2006). The married tend to ignore problems with their spouses because the relationship is an important source of emotional closeness and intimacy (Luong, Charles, & Fingerman, 2011). Recent empirical research however suggests that in marriage, the husband's health has numerous effects on the wife's wellbeing while the wives' health does not have comparable effects on the husband's satisfaction with life (Iveniuk, Waite, Laumann, McClintock, & Tiedt, 2014), Researchers have identified a little relationship between education and life satisfaction. However, the relationship seems to disappear when income and occupation are statistically controlled. In other words, the relationship between education and life satisfaction is probably due to the fact that higher levels of education are related to higher incomes. Cuñado and de Gracia (2012) suggest that education has direct and indirect means of influencing individual well-being. The direct means refers to the increased self-confidence that higher education provides, while the indirect means are related to the opportunities of quality jobs, income, etc. that it provides. Education also appears to be more highly related to life satisfaction for individuals with lower incomes and in poor countries. Probably, poor individuals obtain better life satisfaction from education because the achievement is beyond their expectations. Blanchflower & Oswald (2011) and Helliwell (2008) suggest that education has a positive association with life satisfaction. Diener, Diener, and Diener (1995) found that higher education was a significant predictor of life satisfaction. In investigating demographic data and life satisfaction, it has been reported that higher education tends to lead towards higher life satisfaction. On the contrary however, the well educated are reported to be slightly less satisfied with life in the developed countries (Veenhoven, 1994). This phenomenon is commonly explained in terms of relative deprivation, the well educated expecting more than they get.
Evidently, numerous studies had been previously conducted on the relationship between subjective well-being and demographic factors. However, past findings were not conclusive and besides, only a few of the studies were conducted among Nigerian samples. Against this background, the following research questions are raised: Will age, gender, marital status and educational qualification independently and jointly predict life satisfaction? Will age,gender,marital status and educational qualification interact to predict life satisfaction? One major contribution of this study would be the implication of its findings for the assessment of individuals on life satisfaction.
The following hypotheses would be tested;

International Journal of Social Science Research
ISSN 2327-5510 2017, Vol. 5, No. 1 1) Age, gender, marital status and educational qualification will significantly jointly and independently predict life satisfaction.
2) Age, gender, marital status and educational qualification will significantly interact to predict life satisfaction

Design
The study is a descriptive survey. The independent variables in the study are: age (at 4 levels; 16-19,20-29,30-45,46-60, and above 60), gender (at 2 levels; male & female), marital status (at 5 levels; never married, married, separated, divorced and widowed) and educational qualification (at 4 levels; no formal education, primary education, secondary education and tertiary education).The dependent variable is subjective wellbeing.

Sampling Procedure
A multistage sampling method was adopted for the study. The first stage involves obtaining the list of all Local Government Areas (LGAs) in Ibadan metropolis from the Ministry of Lands and Housing and selecting the five major LGAs from the existing eleven LGAs, using purposive sampling method. The second stage involves obtaining the list of enumeration areas (EAs) for the selected five major LGAs in Ibadan metropolis from the National Population Commission (NPC). The researcher randomly selected 50 EAs i.e. ten EAs from each Local Government Area (LGA) by assigning numbers to the enumeration area (EA) names, calculated the sample fraction, randomly selected the first EA and finally selected every nth on the list for the remaining EAs, based on the sample fraction. Stage three was the point at which the EA maps for the selected areas were obtained from the National Population Commission to determine the number of houses and their locations in the selected EAs in each of the LGAs. The fourth Stage involves random selection of households among the identified houses from each EA by picking and marking every other household or balloting to select a household where there are blocks of flats. The last Stage was the sampling of all heads of households residing in the marked houses.

Participants
220 participants were sampled in each LGA, making a total of 1100 participants. Participants consisted of house-owners and renters drawn from the high, low and medium density areas of the five major Local Government Areas (LGAs). The participants' age range was 42.11+15.20 years. 443 (43.8%) of the participants were males while 569 (56.2%) were females. Participants' educational qualification was 9.9% no formal education, 23.7% primary education, 29.9% secondary education and 36.5 tertiary education. Participants' marital status was 79.6% married, 1.7% separated, 0.27% divorced, 4.9% widowed and 13.5% never married.

Research Instrument
The main instrument for data collection for this research was a structured questionnaire  Neugarten et al. (1961). It has 3 response formats which are disagree, agree and don't know. It measures 5 domains of life such as zest for life, resolution and fortitude, congruence between desired and achieved goals, high physical, psychological and social self-concept, happy and optimistic mood tone. The 20 item has become the most used survey instrument for older adults (Helmes, Goffin & Chrisjohn, 1998). The Cronbach alpha reported ranges from 0.79 to 0.90. Its a three-point scoring system which rates an agree response as 2, I don't know response as 1 and a disagree response as 0. The total score on the scale was based on participant's agreement with specific responses to individual items. Neugarten et al.(1961) obtained a mean score of 12.4 (SD, 4.4).The possible range with one point given for each agreement is 0-20. The Cronbach alpha recorded for this study is 0.55. The mean for this study is 19.61. Participants who scored above the mean were categorized as being satisfied with life while those who scored below the mean were classified as not being satisfied with life.

Research Setting
The study took place in 50 enumeration areas or neighbourhoods across the five major Local Government Areas (LGAs) in Ibadan metropolis, South Western Nigeria.

Procedure for Data Collection
The researcher located the randomly selected enumeration areas or neighbourhoods within the five major Local Government Areas (LGAs) with the help of the seven experienced staff members of the National Population Commission who were the research assistants for this study. Enumeration area maps were used to identify the selected enumeration area boundaries. Having randomly selected all the houses in the selected enumeration areas, households were identified. Having identified the households, the researcher identified heads of households of each of those selected houses and presented the researcher's letter of introduction to them. The research assistants also applied their vast experience in getting the cooperation of some, initially, uncooperative participants. Participants were made to understand that the purpose of the exercise was purely academic and therefore the confidentiality of their responses was guaranteed. The researcher sought their permission to mark their houses with chalk before the commencement of the administration of questionnaires. Having agreed to participate in the study, participants were requested to sign the consent forms before the questionnaires were given to them to test the stated hypotheses. Questionnaires were administered under the condition of anonymity. Some questionnaires were completed and returned immediately, some were collected some hours later, some were collected the next day, some were collected days after while some were never returned. The delay in collection of questionnaires was partly due to the tight schedule of some respondents and the number of items involved in the International Journal of Social Science Research ISSN 2327-5510 2017 scales. Some of the questionnaires that were either not well filled or completed were discarded. Out of the one thousand one hundred questionnaires administered, only one thousand one hundred and twelve were completed and returned. The completed copies were scored and analyzed with Statistical Package for the Social Sciences (SPSS) software.

Statistical Analysis
The statistical tool employed in this study were co relational analysis, moderated hierarchical multiple regression analysis and analysis of variance (ANOVA). The hierarchical moderated multiple regression analysis was used to test the moderating effect of age on the relationship between socio-demographic variables and life satisfaction.

Analysis of Results
To facilitate interpretation of significant interactions, some 5x5 interaction tables and graph were plotted to show the moderation effect, where applicable. ISSN 2327-5510 2017  influence of the predictor variables shows that of age and marital status independently and significantly influenced life satisfaction (β=-0.12, t=-3.50, p<.01; β = -10, t=2.65, p<.05) respectively. However, sex and educational qualification were not significant. Also, the independent influence of the predictor variables shows that age and interaction of age and marital status predicted life satisfaction (β=-0.13, t=-3.58, p<.01; β = -0.08, t=1.98, p<.05) respectively. The moderated relationship between age and marital status contributed significant 0.5% (R 2 = .15, ∆R 2 = .005, F (1, 1004) = 3.39, p<.01) increase in the prediction of life satisfaction and size of the co-efficient of determination in the model. The correction term for the moderated relationship between marital status and life satisfaction was significant (ct=0.08, t=1.98, p<.01). A breakdown of the moderation effect was carried-out using a 5x5interaction graph presented in fig 1: International The graph reveals that younger participants (below 30years) who were divorced or widowed or never married were more satisfied with life than those respondents that were married or separated. Similarly, respondents (between 30-60 years) who were divorced or widowed or never married were more satisfied with life than those that were married or separated. Also, respondents above 60 years of age who were never married were more satisfied with life than those married, divorced, separated or single.

Discussion
The first hypothesis was confirmed. The findings reveal that age, gender, marital status and educational qualification jointly and significantly predicted life satisfaction. This result is consistent with the previous studies that have reported positive association between socio-economic factors and life satisfaction. Sacks, Stevenson and Wolfers (2010) and Aldousari and Kassouf (2010) document a positive relationship between subjective wellbeing and education, income and social relationships. Similarly, Frey (2008) found demographic variables like, income, marriage, education and religiousity to be significantly associated with individual happiness.
The independent influence of the predictor variables reveals that age and marital status International Journal of Social Science Research ISSN 2327-5510 2017 independently and significantly influenced life satisfaction. This outcome is in line with the conclusion of Charles and Carstensen(2009) who argue that individuals experience more life satisfaction as age increases because, as time goes by, older people become aware of mortality and therefore spend more time in activities that contribute to their immediate well-being rather than pursuing goals that are meant for the future. The significant independent influence of marital status on subjective wellbeing is in agreement with the study of Musick & Bumpass,(2012) who found marriage to be linked with adult psychological well-being. In western countries, numerous studies have confirmed that married and cohabiting individuals were more likely to report greater life satisfaction than their unmarried, divorced, and widowed counterparts (Soons, Liefbroer, & Kalmijn, 2009;Musick & Bumpass, 2012).
However, gender and educational qualification did not significantly influence subjective wellbeing. The insignificant outcome of gender influence is consistent with previous findings by Casas, Figuer, Gonzalez, Malo, Alsinet and Subarroca (2007) and Froh, Yurkewicz and Kashdan (2009). They conclude there was no difference in the level of life satisfaction across gender. The insignificant result of educational qualification is in conformity with Salinaz-Jimenez, Aartes and Salinaz-Jimenez (2011) who conclude that the negative relationship between educational attainment and life satisfaction is explained by the tendency of the highly educated individuals to have greater job expectations and personal ambitions than the less educated and if such expectations fail, it may lead to a decline in life satisfaction. Generally, the negative association between education and life satisfaction is usually explained in terms of relative deprivation, the well educated expecting more than they get. The knowledge and exposure that comes with education encourage more ambition and higher achievement targets and subsequently more stress which might lead to dissatisfaction for the highly educated people.
The second hypothesis was also confirmed. The result demonstrates that age interacted with marital status to influence subjective wellbeing. The interaction graph reveals that younger participants who were below the age of 30years, who were divorced or widowed or never married were more satisfied with life than participants who were married or separated. This is consistent with the previous findings by Blanchflower and Oswald (2008) and Steptoe, Deaton & Stone,(2015) who report that life satisfaction reduces with age in lower-income and middle-income countries. Baird, Lucas and Donnellan (2010) contend that there is a longitudinal decrease in life satisfaction from age 70 and above. Gerstorf, Ram, Roecke, Lindenberger and Smith (2008) confirm there is a great reduction in life satisfaction about 2 or 3 years before death. Despite the report that old people possess the ability to adapt and manage their circumstances. such ability often reduces among the very old people. Gwozdz and Sousa-Poza, (2010) note that there is a significant reduction in subjective wellbeing among the oldest old. The implication of this is that subjective wellbeing in old age may only occur in old people, in early old age.
Further analysis also reveals that participants between ages 30 and 60 years, who were divorced or widowed or never married were more satisfied with life than those that were married or separated. Also, participants above 60 years of age who were never married were more satisfied with life than those married, divorced, separated or single. The implication is International Journal of Social Science Research ISSN 2327-5510 2017 that married participants reported lower levels of subjective wellbeing compared to the singles, never married, divorced and widowed. Although several studies have supported the claim that marriage is related to happiness and life satisfaction, many recent studies have reported contrary findings. This result conforms with studies such as; Proulx, Helms, & Buehler, (2007), Dush, Taylor, & Kroeger, (2008), Luhmann, Hofmann, Eid, & Lucas, (2012 and Carr, Freedman, Cornman & Schwarz, (2014).These researchers argue that what makes married people satisfied with life is not the marriage itself but the quality of the marriage. They suggest that a problematic marriage is not capable of guaranteeing happiness. Similarly, Bookwala, (2012) observe that individuals' overall life satisfaction is influenced by their satisfaction with their marriages. Longitudinal studies have also shown that as marital quality decreases, depression increases (Karney, 2001;Proulx & colleagues, 2007). In addition, Gere & Schimmack, (2011) have also identified conflict between partners as a major cause of lower subjective well-being. This argument seems logical, considering the findings of Reblin, Uchino and Smith (2010) which link negative relationships with weak or ineffective social support and a source of stress (Holt-Lunstad, Uchino, Smith, & Hicks, 2007). Evidently, recent studies are in support of marital quality being positively related to subjective wellbeing, however, Bookwala, (2012) and Jackson, Miller, Oka, & Henry, (2014) contend that the relationship between marital quality and subjective wellbeing is usually stronger among women than men.

Conclusion
The findings reveal that age, gender, marital status and educational qualification jointly and significantly predicted subjective wellbeing. This study found no main effect of gender and educational qualification on subjective wellbeing and no significant interaction effect of age/gender and age/educational qualification but there was a significant main effect of age and marital status on subjective wellbeing. The result further demonstrates the interaction effect of age and marital status on subjective wellbeing.

Implication and Recommendation
The result of this study has indicated both the joint predictive strength as well as the the main and interaction effect of age, gender, marital status and educational qualification on subjective wellbeing. This emphasizes the significance of socio-demographic factors in the assessment and eventual improvement of subjective wellbeing. This result has implications for clinical practices especially as it relates to counseling. Some of the socio-demographic characteristics appear to be factors of the environment rather than being innate. They appear to be behaviours that could be learnt and achieved through counseling and reinforcement to individuals lacking in those aspects of life. Socio-demographic factors such as marriage, education, relationships and so on, should be included in counseling.
It is therefore recommended that subjective wellbeing experts and researchers generally, should include numerous socio-demographic factors that could possibly predict subjective well being in order to have a comprehensive assessment of the relationship between socio-demographic characteristics and subjective wellbeing.  ISSN 2327-5510 2017