The article cites longitudinal studies showing randomly selected couples are happier before they have kids and after kids leave the house than while the kids are home. It shows a significant portion of randomly selected parents would rather do chores (e.g., grocery shopping) than spend time with their kids. Neither of those is adequately explained by selection bias.
Selection bias may explain why childless couples, on average, are happier than couples with children. But not any of the dozens of corroborating data points presented in the article.
Much better to measure these things indirectly (without a questionnaire). I don't know how the cited study did it.
It's also safe to assume you'll only be happy for a year after your kids leave, and this likely explains the newlywed syndrome that they're happy for the first year and then reality kicks in.
This study shouldn't be that surprising. It is adequately explained by Fruedian/Jungnian psychological theories.