Innovations in the Measurement of Gender Role Attitudes 2 |
|
| Session Organisers |
Professor Corinna Kleinert (Leibniz Institute for Educational Trajectories (LIfBi) and University of Bamberg) Dr Gundula Zoch (Leibniz Institute for Educational Trajectories (LIfBi)) Dr Stefanie Heyne (LMU Munich, Department of Sociology) |
| Time | Wednesday 17th July, 11:00 - 12:30 |
| Room | D24 |
In large-scale surveys, gender ideologies are commonly measured by item sets which cover attitudes towards different gender roles in various spheres of life. They aim at measuring individual support for various separate spheres such as women’s labour market participation or the division of housework (Davis & Greenstein 2009). Beliefs in gendered spheres are referred to as traditional ideologies, while egalitarian gender ideologies describe beliefs that value an equal division of paid and domestic work among men and women as well as equal participation in different occupational fields and roles.
The gender role attitude items used in survey research so far differ with regard to the extent of gender traditionalism expressed as well as with regard to the spheres of life covered. Comparative research has found gender role attitude items slanted towards traditionalism to produce clearer differences between countries than more egalitarian items. However, these items seem less suitable to capture ideology change over time in groups with relatively egalitarian views. Conversely, items which capture more egalitarian beliefs might lead to insufficient discrimination between traditional and non-traditional respondents (e.g. Braun 2008). Moreover, previous research has argued that due to the overall decline in more traditional ideologies in society, gender ideologies do not comfortably fall onto a continuum of traditional and egalitarian gender role attitudes anymore, but are increasingly multidimensional in nature. Thus, commonly used item sets might not be sufficient anymore to capture traditional and egalitarian ideologies as well as different types of egalitarianism.
For this session, we welcome contributions that
- investigate problems of measuring gender role attitudes, either as determinants or outcomes, for example of life-course events or policy changes, by focusing on single-country or comparative studies,
- introduce new and innovative conceptual frameworks which are measurable in survey research,
- provide evidence from new measurement approaches based on a broad range of data types and instruments, i.e., small- or large-scale, qualitative and quantitative studies, longitudinal studies, experiments, vignette studies or mixed method studies.
The stream particularly welcomes contributions that aim at developing a more valid measurement of gender role attitudes which is suitable for investigating ideology changes over time as well as intergenerational ideology transmission in large-scale surveys with limited time and heterogeneous respondents.
Keywords: gender role attitudes, measurement, measurement error, item development
Dr Florian Schulz (State Institute for Family Research at the University of Bamberg) - Presenting Author
Attitudes towards sharing housework in couples are an elementary component of gender ideologies, as the notion of women as homemakers is inseparably knotted to the traditional family model. There is an empirical research tradition on attitudes towards housework which largely relied on item-based measurement approaches of attitudes. This situation is problematic because, first, most empirical studies in the field of housework have been working with either general gender ideologies or different dimensions of housework, for example enjoyment or standards, and not in particular with housework sharing. Second, clear interpretation of “agree-disagree-items” has been problematic in previous research, as it usually tells us little about the division of labor which women and men actually favor. I propose an innovative factorial survey approach to measure the attitudes towards housework sharing. This vignette approach is said to be advantageous compared to item-based opinion surveys when it comes to studying respondents’ attitudes. Human judgments and their determinants, even in cases of sensitive issues, can be addressed more directly without the need for deducing them indirectly from the assessment of statements and with a lower risk of social desirability bias. Further, the approach offers the possibility to control for individual as well as contextual factors when studying the determinants of the outcomes. As an empirical application, I present data from a factorial survey on housework sharing of 1,200 German women and men, in which I have measured attitudes towards female-typed chores (cooking, cleaning, laundry), male-typed chores (repairs, car, paperwork) and housework in general (generic term of housework). Empirical results showed that attitudes towards female-typed tasks are clearly distinct from attitudes towards male-typed tasks but not from housework in general. Interpreted with caution, this might indicate that women and men have female-typed routine housework in mind when thinking of and being surveyed about housework in general.
Ms Sabine Düval (University of Munich) - Presenting Author
Professor Katrin Auspurg (University of Munich)
Measures of gender role attitudes are frequently used to explain social inequalities such as differences in fertility, gender specific divisions of labour or labour market inequalities. Most measures are based on standard items on respondents’ attitudes towards gender roles, often with an emphasis on the division of housework.
Those items have been criticized because of their one-sided focus on the role of men versus women (e.g. Braun 2008). But there is another problem which hasn’t received much attention: The items contain only very few information on the “men” and “women” to evaluate. We suspect (relying e.g. on theories of statistical discrimination) that respondents impute missing information in a gender-specific way, based on typical arrangements in couples. If this is true, gender role items would not only measure gender ideologies, but also reflect (group specific) beliefs about consisting (gender) inequalities. For instance, respondents might only agree that women should do a larger share of housework than men because they assume that women provide less market income or have more time due to fewer labour market hours. If this is true, the traditional item questions would overestimate traditional attitudes, since also respondents who are only in favour of a gender neutral, equal exchange would agree with those “traditional” items.
To test our assumptions, we implemented a factorial survey experiment in wave 10 of the German Family Panel (pairfam). About 1300 respondents evaluated 3700 vignettes on normative attitudes about an adequate division of housework. The vignettes contained different amounts of information on labour market hours and the relative earnings of both partners. First results suggest that attitudes about an adequate sharing of (house-)work depend indeed on the (amount of) available information. Based on our results it can be questioned if previous studies overstated the impact of gender roles due to invalid measures.
Dr Jessica Walter (GESIS Leibniz Institute for the Social Sciences) - Presenting Author
Measures of gender role attitudes are used in many national and international surveys. Most of these measures were developed in the 70s and 80s and focus on the traditional male breadwinner model. However, social developments of the last decades such as increasing female employment or an increase in higher education challenge the assumption that the measures are still adequately since the male breadwinner model lost importance within many societies. After reviewing the measures in light of the social developments, I conclude that measures of gender role attitudes need an update. Using the example of the German General Social Survey I present a revised measure of gender role attitudes according to the social developments. It comprises two factors – a traditional and a modern factor – represented with nine items. Previously asked traditional items are supplemented with items for the evaluation of the role of the father in the family and more egalitarian models of division of labor in the family. Nevertheless, the measure has been designed to ensure analyses over time. This presentation will provide an overview of how this measure was developed and implemented and how it improves the measurement of gender role attitudes for social science research.