Trinkets is a section of the Voices of Equality blog. I am Debora Diniz and I am here with you every Saturday. Today's section is called Tools, and is a conversation about tools, forms, and ways to collect data, particularly in social research.
Here I tell a story about how we decided not to include a question about religion in the ballot box in the questionnaire that we collected the magnitude of abortion in Brazil. The 2010 National Abortion Survey (PNA), covered all urban Brazil. We entered women's homes, between 18 and 39 years-old, with a ballot with five questions.
At first, we had six questions. The last of them was: what is your religion? On the top, we had the question if she ever had made an abortion.
What in the pre-test phase, known as instrument validation, what we discovered was that the woman kept on answering the questions. When she got to the last one, "what is your religion," she changed the answer about whether she had ever made an abortion or not. The answer about religion fired a moral filter on the truth of her experiences on abortion.
As a result, we removed the question on religion of the ballot box and put it in the open questionnaire, that was paired to the one for the ballot box. The woman answered the five questions, deposited in the ballot box and then we did an open questionnaire, where, among other things, was the question about her religion. What does this shows us?
The importance of thinking not only what to ask, how to ask, but also the order of what to ask. You see that such a simple question as this would change the quality of the answers on the magnitude of abortion in Brazil. The Tools series has this goal.
We exchange experiences. If you have other stories to tell us, or have any questions on gathering research instruments, what is the best way to get to a particular group, write to us. I will be delighted to come back here and continue this conversation.
See you soon.
