Relationship Between Divorce and Family Models | Proposal
Aim and Hypothesis
“The conventional nuclear family is already a thing of the past,” writes Madeleine Bunting, author of ‘Willing Slaves: How the Overwork Culture is Ruling Our Lives[1]’. Current statistics do indeed paint a worrying picture; in the past 30 years the rate of divorce has doubled[2], Britain now having the highest divorce rate in Europe. Concurrently, the proportion of women in full-time employment has trebled, with maternal employment having increased by 8% to 65%[3]. As a firm believer in the Nuclear Family model, these findings and predictions are of particular personal concern. For this reason I have chosen this area as the focus for my sociological research. I have chosen to concentrate my study on the relationship between the ever-changing female sex stereotype, and the heightened rates of divorce. My test hypothesis is therefore as follows:
“Divorce is more prevalent amongst families where the mother does not willingly accept the traditional female sex employment stereotype.’
Contexts and Concepts:
The context of this hypothesis is best understood in light of two research studies which have been conducted previously in this area. The first was an exploratory research study conducted by Geoff Dench of Middlesex University into men’s family roles, entitled “The Place of Men in Changing Family Cultures.[4]” In this study, Dench identified two main family types; the ‘conventional family’ and the ‘deregulated family.’
The conventional family model represents a traditional family formulation in which each member has interpersonal rights and obligations arising from their marriage [or birth, if children]. The reciprocal support structure inherent in this model results in individual free choice being limited; purely personal interests may therefore need to be compromised. Dench identified that members of this group would be more inclined to see a distinction between what may be good for them as an individual, and what may be good for the family as a whole.
The deregulated family model represents an individualist family formulation in which each member is responsible for upholding their own rights and negotiations of those rights. To this extent, individual choice is prevalent within this group; autonomy being the over-arching ideal. Respondents in this group expressed a view that traditional gender roles were a source of social injustice, and rejected the notion of the ‘conventional family’ as ideal on this ground. Dench identified that supporters of this model tended to be young, childless couples.
Whilst this study yielded several very interesting findings, it is one conclusion in particular which is most relevant to the study of this paper: Dench discovered that the ‘highest levels of personal and marital satisfaction were among people who lived in traditional families. The highest levels were connected to families in which the man worked and the woman was at home. The lowest levels occurred when the woman worked and the man stayed at home.’
This is the very conclusion which I am attempting to prove within this research study, via my hypothesis that “Divorce is more prevalent amongst families where the mother does not willingly accept the traditional female sex employment stereotype.” The concept of ‘traditional female sex employment stereotype’ describing the situation where the role of the husband is as breadwinner of the family [i.e. provides the primary family income], and where the role of the wife is as ‘housewife’ [i.e. the primary domestic labourer].
The second research study is entitled “The growth of Lone Parenthood: Diversity and Dynamics[5]”, and was conducted by Karen Rowlingson, Stephen McKay and Richard Berthoud on behalf of the University of Derby’s Economic and Social Research Council. Their findings, whilst not specifically concerned with the issue of the female sex employment stereotype, suggested that divorce occurs more frequently in families where the woman leaves her full-time employment to look after a newly arrived child.
Prima facie, this finding would suggest that my conclusion is flawed; after all, surely a woman who leaves work to look after her newly born child is adhering to the traditional employment stereotype, and should therefore be less likely to engage in divorce or separation from her partner. This finding however does not contradict my hypothesis, rather helps to contextualise it; this finding from the study concerned women who had previously been non-traditional employment stereotype conformists, but had been forced to give up their work and look after their child through circumstance. A woman who does not willingly accept her traditional role is of course more unlikely to be unhappy, as her ‘individual autonomy’, which as discussed by Dench is held as the over-arching ideal by a subscriber of the non-traditionalist model, is not being allowed to predominate. This helps to explain why my hypothesis is limited to/focussed upon women who do not ‘willingly accept’ the traditional gender employment stereotype.
Main Research Method and Reasons:
In light of the obvious difficulties with actually asking random adults about their marital status and employment, I have decided that the best way to glean this information is by asking my fellow students at school to complete a survey questionnaire [a method usually associated with the positivist school of sociology]. It is of course necessary that a significant proportion of my sample population have parents who are divorced, and visa versa. The problem with this requirement is that a random sample of students may not throw up sufficient of the former category. What I therefore propose is a one day trip to Somerset House in London, and armed with a complete list of all students at my college, I will cross reference their surnames against the Divorce Register[6]. Once I have identified all the students whose parents are now divorced, I will randomly select a proportion of these students to take part in my survey by assigning each student with a number and using a random-number generator program on my computer to decide which students will be selected. By the same random-sampling technique, I will select an identical number of students whose parents are not divorced.
The benefit of conducting this prior research to identify students whose parents are divorced is that it means that the questionnaire itself does not need to ask these people whether or not their parents are separated. To ask this question may prove to upset the students; after all the divorce of their parents is not something which most students would wish to discuss or even recall during a sociological study such as mine. For the student’s whose parents are apparently still together, married, it is important to confirm that the student still lives with both parents; after all just because the parents of these student’s do not appear on the Divorce registry doesn’t mean they are still living together- they could be separated without divorce, or in the process of applying for a decree nisi/absolute.
The questionnaire itself will ask questions designed to glean the following information:
From those students whose parents are now divorced:
If the student’s mother, for the majority of the student’s life, was engaged in full-time employment prior to separation.
If the students mother, for the majority of the student’s life, was engaged in part-time employment prior to separation.
If the students mother, for the majority of the student’s life, was a ‘housewife’ prior to separation.
The reasons for the above answers, if known. [e.g. my mother always though that it was important to look after the family while the man went out to work etc.]
From those students whose parents are still married, and living together:
If the students mother has been employed on a full-time basis for the majority of the student’s life.
If the students mother has been employed on a full-time basis for the majority of the student’s life.
If the student’s mother has been a ‘housewife’ for the majority of the student’s life.
The reasons for the above answers, if known. [e.g. my mother always though that it was important to look after the family while the man went out to work etc.]
In light of the fact that I will need to compare all the data if I am going to test my hypothesis, I plan to use the above information to calculate a total ‘employment score index’ for each student sample. One simple way to achieve this is by structuring the questions of the survey as score lines, i.e. On a scale of 1-9 was your mother, for the majority of your childhood, 1= fully employed 2= fully employed but took some time of work to look after me in my early years 3= was fully employed for some of the time, but also spent a good proportion of those years at home as housewife… 9= has always been a housewife since I was born. In this way the indexes of each student sample whose parents are not separated can be compared with the indexes of each student sample whose parents are divorced, a comparison from which a trend may or may not be identified, and may or may not be in line with my hypothesis.
2 pre-tests will need to be conducted by a sample of 25 students in order to tweak the questionnaire[7], so that we can be assured that the final questionnaire is appropriately constructed.
This survey will create qualitative results; the limited size of the sample study precludes there being any reliable quantitative conclusions, and besides, there are clearly other factors which cause couples to separate, and therefore from such a study, we can only sensibly be interested in identifying a qualitative trend.
Potential Problems:
One problem associated with proposed study is the fact that due to the time restraints in which I would be operating, the limited size of my sample frame would perhaps not yield sufficient results from which a reliable trend/conclusion could be drawn. This problem could of course be rectified without changing the method, but rather by increasing the sample frame, and as such is not what could be described as an inherent problem.
Another problem is the fact that we are questioning students about their mothers employment history throughout their childhood. Whilst some students may remember these details with great clarity, it may well be the case that certain students are unable to recall this information with sufficient accuracy. One solution to this problem, and a way of avoiding erroneous results through sample guesswork, would be to ask each student on a scale of 1-5 how accurate they believe their assessment to be.
Finally, it may also be the case that there are simply not enough students from divorced families in my school to provide sufficient and therefore reliable data for this side of the comparison. If this is the case, it may be essential to extend the survey across several additional local colleges, although the time-frame available would not make the necessary research for this sample frame extension feasible.
When asking the students to confirm that their parents are still married and living together, students may not be willing to admit that their parents are currently in the process of divorce or separated, as this is often perceived to be shameful by the children of such parents. This will mean that data from students whose parents are in fact separated [but not divorced] will be erroneously attributed to the data for students whose parents are still married and living together, thus causing inaccuracies to any trend which is identified. One solution to this problem would be to allow each sample student to conduct the survey in the privacy of their own home etc., in this way there is no reason to lie as no-one will ever discover their identity.
