There are three forms of Gender Bias, which can occur simultaneously but are separate problems.
Androcentrism means that problems and perspectives are implicitly investigated in research which predominantly affect men although the findings are generalized to cover all people. This bias takes three sub-forms:
• Over-generalization occurs when one gender is excluded from the data collection process. If conclusions are drawn on the basis of such data, this amounts to an inadmissible generalization of the data that has been collected. If, for example, medications are tested only on young men, the findings regarding effects and dosage may possibly be incorrect or even harmful for women.
• Men as norm. Many research projects are based on norms oriented to men, against which women are then “measured”. In these cases, the different realities of women’s lives can only be depicted as more or less large deviations from the norm. One example is full-time working without interruptions being declared to be the “normal employment situation”.
• Paradoxical gynocentrism is a special aspect of androcentrism. Men are excluded from areas in studies that are deemed to be “typically female”, such as family and private household, caring and reproduction. This is the case, for example, when studies on the situation of single parents are restricted just to women.
2. Gender insensitivity
Gender insensitivity is present if biological sex or social gender is ignored as a variable. This phenomenon is also described as gender blindness. The following three sub-forms can be observed:
• Familialism: “Household” or “parents” are used as the smallest unit for analysis. Research-relevant information affecting individual members of the household or family differently according to their gender are thus neglected.
• Decontextualization: No attention is paid to the fact that similar situations can have different effects for genders. For example, questions as to the compatibility of career and family affect mothers and fathers in different ways. Conventional norms require that women decide in favor of child-rearing and against paid employment, while men are assigned the role of family breadwinner.
• Assumption of equality of women and men: Assumptions of equality in areas in which it may not exist. Women and men are, however, affected to different degrees by restrictions existed in the sphere of employment because of unpaid work. Analyses of the employment situation mostly focus, however, on integration into employment and thus ignore unpaid work (housework, childcare and nursing caring). If a gender-differentiated comparison of participation in employment fails to take account of this, the research findings will be biased as far as gender equality is concerned and thus of little value.
3. Double standards in evaluation (open and concealed)
Double standards in evaluation distort research findings when similar or identical characteristics or behaviors of women and men are evaluated or investigated in different ways. Here are two typical sub-forms:
• Gender dichotomies occur when genders are treated as two completely separate groups. Scales attributing for example various characteristics as typically female and typically male ignore the fact that most personality characteristics occur in all human beings to a greater or lesser degree. Differences between women and men are exaggerated, while common features are at the same time overlooked.
• Gender stereotypes attribute natural characteristics to women and men and do not understand that these are socially-attributed expectations of male and female gender roles. Research findings may also be distorted by double standards if the same behavior is differently interpreted and evaluated because of attributions resulting from gender stereotypes. This is the case, for example, when the statement by a woman that she likes cooking and looking after children is seen as typical female behavior and accordingly given a positive evaluation while the same statement by a man is interpreted as deviating behavior and given a correspondingly negative evaluation.
Berliner Zentrum Public Health (Hg.): Zu mehr Gleichberechtigung zwischen den Geschlechtern: Erkennen und Vermeiden von Gender Bias in der Gesundheitsforschung, Berlin 2002.
Wroblewski, Angela/Leitner, Andrea: Benchmarking Chancengleichheit: Österreich im EU-Vergleich, Institut für Höhere Studien Wien, Reihe Soziologie 67, Wien 2004
Eichler, Margrit/ Fuchs, Judith/ Maschewski-Schneider, Ulrike: Richtlinien zur Vermeidung von Gender Bias in der Gesundheitsforschung, Zeitschrift für Gesundheitswissenschaften, 8. Jg. 2000, Heft 4, S. 293-310.
Margrit Eichler: Nonsexist Research Methods - A Practical Guide, 1988.