Take the Dads4Daughters Test
Split Second Research sponsors the Dads4Daughters campaign in collaboration with Blinc Partnership and St Paul’s Girls’ School.
Take the Dads4Daughters Test – How gender biased are you? Click the logo below:
The test has been featured in:
If your company took part in the test and you would like to have your company’s results, please contact us via email@example.com or +44 (0) 7878455944
The Dads4Daughters Test is based on a commercial test designed to measure attitudes towards brands, TV adverts, and other marketing materials. These attitudes are measured implicitly, that is, they are inferred from reaction times to words images presented on the screen. The test bypasses the need to ask explicitly about someone’s views or attitudes. This is important because often what people say they will do or what they tell you about how they feel is often at odds with how they behave! Furthermore, in difficult issues such as sexism or racism, people may be reluctant to tell you how they truly feel and in some cases they may even hold certain attitudes that they are unaware of until they are provoked.
The commercial test is itself based on the evaluative priming paradigm in academic research (e.g., Fazio, et al., 1986)1. The first phase of the test is to detect target emotion words as belonging to either one category (e.g., Happy) or another (e.g., Sad). In the second phase, the task is the same but the target emotion words are preceded very briefly by ‘primes’. These primes are either congruent with the target word (the prime is Joy when the target is Happy, or the prime is Gloomy when the target is Sad) or incongruent (the prime is Gloomy when the target is Happy, or the prime is Joy when the target is Sad). The task can be performed quicker and with fewer errors when the prime and the target are congruent than when the prime and the target are incongruent.
In the Dads4Daughters version of the test, the targets were female and male words, such as She and He. Primes were 24 words related to professions, roles, personal qualities, or career fields (e.g., engineering, manager, leader, and so on). Trials are divided into female trials (where the target is a female word that invites a specific response, e.g., press D on the keyboard or swipe left) and male trials (where the target is a male word that invites a different response, e.g., press K on the keyboard or swipe right). The logic is: if the test-taker subconsciously associates a career field (e.g., engineering) as being male, then they will be quicker to detect the male target on ‘male’ trials than the female target on ‘female’ trials when the prime is engineering.
About 10,000 people have taken the test and from all walks of life (from bus drivers to CEOs).
Note that the test does not require an explicit evaluative judgement, there is always a correct answer on each trial. Also, people often think they can out-game the test – but in fact there is no way to ‘play’ the system because the task is always the same – it is a test that is difficult to fake. The way the test measures an attitude is not to do with accuracy or generally how fast the response is, but through a comparison of reaction times. So, the association between say, engineering, and the concept Female or Male is detected (or implied) by differences in reaction times to detect the female and male targets when they are both preceded by the same prime (engineering).
For each test-taker, the result is a measure of adherence to traditional views of gender roles – that some roles are associated with men and others with women. The more a test-taker associates skilled professions, or more senior roles, or more desirable personal qualities with men and not with women, then the stronger is their measure of gender bias.
If you would like your employees to take the test and you would like to know how they compare with other companies, please contact us via firstname.lastname@example.org
1Fazio, R. H., Sanbonmatsu, D. M., Powell, M. C., & Kardes, F. R. (1986). On the automatic activation of attitudes. Journal of Personality and Social Psychology, 50, 229–238. doi:10.1037/0022-3518.104.22.168