A new study suggests that the cliché of a full-time college student working a low-wage job to pay her tuition and getting lower grades than she’d have if she wasn't working is more fiction than fact.
If the student works fewer than 20 hours a week, she may, in fact, have a higher grade point average than her jobless peers and be spending her paychecks on “beer money” or other non-tuition expenses.
These are findings outlined in “Parental Transfers, Student Achievement and the Labor Supply of College Students,” forthcoming in the Journal of Population Economics, by Charlene Kalenkoski, an associate professor of economics at Ohio University, and Sabrina Wulff Pabilonia, a research economist in the Bureau of Labor Statistics’ Division of Productivity Research and Program Development. The two economists wanted to learn how work affects students’ academic performance and what might motivate them to take on more hours of work.
http://www.insidehighered.com/news/2009/10/08/work
Kalenkoski has an interesting paper with Foster on the debate as to whether it is more appropriate to fit censored regression (Tobit) models using maximum likelihood estimation or linear models using Ordinary Least Squares (OLS) --- to explain individuals’ allocations of time to different activities as recorded in time-diary data.
ReplyDeleteOne side argues that estimation of Tobit models addresses the significant censoring (i.e., large numbers of zeros) typically found in time-diary data. The opposing side argues that reported zeros represent measurement error rather than true non-participation in the activity, in which case OLS is preferred.
http://oak.cats.ohiou.edu/
~kalenkos/Foster_eIJTUR.pdf
The Mostly Harmless econometrics book also has a discussion. They don't Tobit or Probit, anything thats not linear as far as I can see.
ReplyDeleteBut when it comes to time allocation aren't there genuine zero's?
I really didn't fly a plane this week, engage in regicide etc. And lots of people didn't work either..
Kevin,
ReplyDeleteI agree that there are some genuine zeroes in time allocation. Furthermore, a good deal of these should be reported correctly when one asks a survey respondent about their "average" week.