People are very quick to make judgements, especially in our current society which seems to favour outrage as a response. This happens especially quickly when it comes to topics like sexism, racism and so on. These are all real problems which affect many people but we must be wary about jumping to conclusions without sufficient evidence.
There is an article in Slate magazine which talks about the Red Cross and racism. The starting point of that discussion was a 2014 safety campaign poster which was pulled last year for being “super-racist.” It earned that dubious distinction because:
A “cool” blonde girl waits her turn by the diving board, for example, and a “cool” fair-skinned dad minds his small child. The vast majority of the “not cool” rule breakers, meanwhile, have brown skin: One boy runs through a puddle, another dives too close to a swimmer, and a little black girl pushes a white girl into the pool.
But then I started wondering if the poster really was racist. At first glance it looks like it could be but are there ways to make sure? There actually are! Trying to find out whether groups are evenly distributed or not is something that scientists have to do all the time. To make sure that what we see is really a true difference and not just something random, we use statistical tests.
One of these tests is the Fisher’s exact test, named after its inventor Ronald Fisher. This test allows you to check whether groups are distributed evenly or not, even when you only have a few members in each group. What’s really nice about these tests is that they don’t care what you are testing; it doesn’t have to be something super scientific. We can even use this test to check whether blacks and whites are evenly distributed in a swimming poster.
To test the poster, I’m just going to use an easy web-based calculator provided by Social Science Statistics. You choose your categories, I will say “white” and “black,” as well as your groups. For the groups I am going with “good” for people following the rules and “bad” for people not following the rules. Categorising every character on the poster, gives me a table which looks like this.
|White||Black||Marginal Row Totals|
|Marginal Column Totals||11||8||19 (Grand Total)|
The Fisher’s exact test will then look at how many white vs black characters there are and how many good and bad characters there are and then determine how likely it is that they would be sorted in such a manner if there were no bias. Just because there is no bias does not mean that everything will be perfectly even. You know that any particular side of a die has a 1 in 6 chance of showing up but, if you roll a die six times, you will not necessarily see every side.
Running the Fisher’s exact test on the table above gives you a p-value of 0,32. The p-value tells you how likely you are to get that distribution at least that biased if the groups and categories are really independent. What it says here is that if you took this poster and randomly recoloured every character so there were still the same number of white and black characters, you would get something this “racist” 1/3 times.
Traditionally, most scientists use a p-value of 0,05 to determine when something is “significant.” If something like this poster came up in an experiment, no scientist would consider it to be anything more than a random distribution with no bias. Especially when considering that half of the lifeguards are not white, there is no evidence that this poster is racist.
That does not mean that it is definitely not racist or that the Red Cross does not have racist policies. I can’t say for sure what was going through the minds of the person or people who made the poster nor can I say anything about the Red Cross but the story started with one woman complaining that the poster was racist because of the distribution of white and brown characters. Statistically, given the size of the groups, there is no evidence of racism in this poster.
Even if the poster is not racist, that does not necessarily mean it is not problematic. There could be undertones even in a non-racist message which do not help a situation where racism is prevalent as seen in the quote below. But, even in such a situation, it is necessary to draw a distinction between racism and a message which can be interpreted badly.
In connection with the lack of images showing African Americans excelling in swimming, the poster doesn’t make you feel welcome — it suggests to a black child that you’re not welcome here.
While challenging racism and other forms of discrimination is important we should guard against jumping to conclusions too quickly. It is necessary to consider the evidence behind our statements and to also consider alternative explanations. It is also the responsibility of the media to analyse and fact check claims rather than just repeating accusations. In some cases, that should extend to considering the probability of different outcomes and whether expectations, for example of perfectly even distributions, are realistic and how much randomness can be tolerated.