March 12, 2013
– Comments (4) |
RELATED TICKERS: GE
Avoid Depression: Don't Look at the Market Today
Apply a little common sense.
If something as insignificant as a market decline sends someone to the hospital, it was just a matter of time before they were headed there anyways.
Also, the 0.6% increase in mental health hospitalizations they claim to be statistically significant, is clearly not.
Based on what? Your intuition or your "into wishing"?
If you're familiar with the FDA approval process, a 0.6% increase is not real world significant in the context of medical outcomes.
Think of all the variables which the study is probably ignoring: weather, seasonality, income, age, gender, insurance, etc. 0.6% is so small that it will almost certainly disappear when controlling for other variables.
Instead of speculating about "all the variables the study is probably ignoring", I recommend you take a look at it. If you think serious academics are completely oblivious to these type of effects, you're very much mistaken -- dealing with them is often at the crux of their work."Finally, the vector of controls in equation (1) accounts for the fact that hospital admissions exhibit strong temporal patterns, both within and across years. All of our main results include year fixed effects to account for long-run changes in health conditions, reimbursements, or other secular changes in population health. Month fixed effects account for seasonality; accidents, for example, are more common in the summer, whereas infections tend to cluster in cooler months. Day of the week fixed effects account for any intraweek variation in admissions. Finally, we include indicator variables for the three days surrounding each of the following holiday periods: New Years Day, 4th of July (Independence Day), Labor Day, Thanksgiving, and Christmas."