There exists a perception (stereotype?) of our online community as somewhat shielded from the problems of the real world. Yesterday I posted a list of questions along the lines of “have you ever been racially profiled?” or “have you been evicted?” This list, by no means exhaustive, is after the jump, and I got enough numerical responses to draw some conclusions which I share here.
All the questions referred to an adverse real-life situation I would not normally expect a person of privilege to encounter. A positive answer to each question for the respondent increased the respondent’s score by 1 from zero. If the respondent had not been in the adverse situation described by the question, but had a close friend or relative who had been, I asked them to increase the score by 0.5. The maximum score was 34 (this is what you would get if you answered “yes” to every question for yourself).
I received a total of 19 numerical or quantifiable responses, ranging from 1 to 20.5. The histogram shows the distribution of the answers, which looks vaguely bimodal, with peaks in bins centered around 2.5 and 12.5. The naive interpretation is that we have an overlap of two groups, one of fairly privileged people and the other of people who encountered a lot more adversity. However, if you run a normality test (Shapiro-Wilk) on the same dataset, it is statistically impossible to say that the scores are not coming from a large population that is normally distributed (and it has a single peak). Were that the case, the 95% confidence interval for the average adversity score would be 8.7±1.9, meaning some (but not great) exposure to bad things in life.
Now let me explain why all of this could be rubbish. People self-selected for the poll, meaning that the sample is likely non-representative. Same applies to any online poll or questionnaire. It is also noteworthy that the question list had to be expanded to include adverse situations relating to LGBTQIA+ experiences, domestic violence, etc., showing that the original version had reflected my own blind spots in these areas, so while the exercise may have been of dubious statistical value, I still find it quite useful personally. Thanks to everybody who participated.