2013
DOI: 10.3233/jem-130376
|View full text |Cite
|
Sign up to set email alerts
|

Tobit or not Tobit?

Abstract: Time-diary surveys collect detailed information about individuals' activities over a short period of time, typically one day. Thus, it is common to see zero time spent in many activities, even for individuals who regularly do the activity. Because of the large number of zeros, Tobit would seem to be the natural approach. However, once it is recognized that these zeros arise not from censoring, but from a mismatch between the reference period of the data (the diary day) and the period of interest (typically muc… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
1
1
1

Citation Types

3
129
0
6

Year Published

2013
2013
2020
2020

Publication Types

Select...
7
3

Relationship

0
10

Authors

Journals

citations
Cited by 197 publications
(138 citation statements)
references
References 0 publications
3
129
0
6
Order By: Relevance
“…Some argue that Tobit models are more appropriate in these circumstances, assuming a latent propensity to do an activity, and that negative values of this propensity are artificially censored at zero in the measure (see for example Sousa-Poza et al, 2001). Others counter that time spent in an activity cannot take values less than zero, that zero values in the dependent variable reflect actual and natural nonparticipation, and that OLS is therefore preferable to Tobit (see for example (Stewart, 2009). We follow this approach.…”
Section: Analysis Planmentioning
confidence: 97%
“…Some argue that Tobit models are more appropriate in these circumstances, assuming a latent propensity to do an activity, and that negative values of this propensity are artificially censored at zero in the measure (see for example Sousa-Poza et al, 2001). Others counter that time spent in an activity cannot take values less than zero, that zero values in the dependent variable reflect actual and natural nonparticipation, and that OLS is therefore preferable to Tobit (see for example (Stewart, 2009). We follow this approach.…”
Section: Analysis Planmentioning
confidence: 97%
“…Unlike in the case of television watching and socializing with others (where there are very few zeroes – no one doing them), physical activity and meal preparation activities are less commonly reported on the dairy day. Because estimates from two-part models are less biased than estimates from Tobit models (Stewart 2013; Daunfeldt and Hellstrom 2007), we use logit models to estimate diary day participation in physical activity and meal preparation and OLS models to estimate minutes spent in these activities (conditional on participation). Our decision to analyze age and gender subgroups is generally supported by model-level Chow tests showing significant differences in OLS age- and gender-specific models and significant age and gender coefficients in pooled non-linear models (available upon request).…”
Section: Data and Proceduresmentioning
confidence: 99%
“…For these reasons, OLS is argued to be preferable to Tobit (see for exampleStewart 2009;Brown and Dunn 2011).…”
mentioning
confidence: 98%