If you read the full article, he maintains that "The substance of my argument that marriage is generally better for men than women remains." One aspect of the survey was misinterpreted, but it didn't actually change the results. The article is about how we need to fact-check data before we buy into it because even researchers make mistakes. However, his mistake didn't change the actual outcome of his study, only one aspect of the data.