Download presentation
Presentation is loading. Please wait.
Published byAshlynn Bridges Modified over 9 years ago
1
Dissemination and interpretation of time use data Social and Housing Statistics Section United Nations Statistics Division Time Use Statistics workshop for Arabic speaking countries, Amman,25-28 April 2011
2
Dissemination and interpretation of time use data Stiglitz commission on the Measurement of Economic Performance and Social progress Aim 1: Identify the limits of GDP as an indicator of economic performance and social progress Aim 2: Consider additional information required for the production of a more relevant picture
3
Dissemination and interpretation of time use data The 2008 report recommends to take into consideration unpaid activities and more precisely “household production” Revival of interest for Time use surveys beyond the traditional concern about labor-leisure tradeoff Time use survey for use in public policy to deal with a large range of social issues (quality of life, gender, work…) Dissemination and interpretations stages are crucial because they are not regular surveys
4
Dissemination and interpretation of time use data 1)Modes of dissemination 2)Issues in dissemination of time use data 3)Examples of processing and interpreting time use data Some key lay-outs from a study carried out based on last French time use survey
5
Modes of dissemination Up to the statistical office to assess the suitability of the differing modes of dissemination Microdata Macrodata Metadata Suitable combinations of formats and media which meet the differing capabilities of users Ex: Eurostat
6
Disclosure control Disclosure control =measures taken to protect statistical data in such a way as not to violate confidentiality requirements as prescribed or legislated Suppression of cells values on the basis of a “sensitivity”criterion Table redesign Perturbing data through the addition of noise
7
Examples of processing and interpreting Introduce a study carried out with some other former colleagues of INSEE Bringing out how poor people use their time in France: context of “Inactivity Trap” Not an exhaustive overview of what can be done but examples of different ways of exploiting time use data
8
Examples of processing and interpreting Descriptive statistics Timing diagrams Econometrics tools Optimal matching
9
Examples of processing and interpreting Descriptive statistics Timing diagrams Econometrics tools Optimal matching
10
Descriptive statistics At the first stage, the statistician can lay out descriptive statistics: On the fact of practicing or not one or some activities On the duration of practicing one or some activities
11
Descriptive statistics
12
Examples of processing and interpreting Descriptive statistics Timing diagrams Econometrics tools Optimal matching
13
Timing diagrams People might be interested in having a dynamic perspective For that, the statistician can set up timing diagrams Timing diagrams represent the proportion of people practicing an activity for each hour around the clock
14
Timing diagrams
15
Examples of processing and interpreting Descriptive statistics Timing diagrams Econometrics tools Optimal matching
16
Econometric tools Descriptive statistics are not sufficient if you want to work “all else equal” Given the complexity of time use survey sampling, it is sometimes required to investigate more complicated modeling. The sampling and the social inquiries often induce biases
17
Econometric tools In our study, regression of duration of practicing an activity on the poverty status by OLS. However the estimations are biased Time dedicated to an activity available providing that the respondent did practice it on the sampled day Actually, the duration of practicing an activity is a censored variable Tobit model
18
Econometric tools 2 nd equation (D): fact of practicing or not a specific activity 1 st equation (Yi): duration of practicing this activity Instrument variable
19
Econometric tools
20
Examples of processing and interpreting Descriptive statistics Timing diagrams Econometrics tools Optimal matching
21
Comparing sequences of activities between all the respondents Coming up with homogeneous groups which share similarities in their use of time and representing their “typical” daily schedule 2 stages
22
1 st stage Computes a distance between every two sequences. All the possibilities to convert a sequence to the other via three operations: suppression, substitution or insertion Each operation is associated with a cost Ends up selecting the minimum general cost as the distance
23
2 nd stage Classification of the sequences: the statistician has to choose the most relevant number of groups to describe the heterogeneity of the population.
24
Graphics
25
Conclusion Crucial topic: should be considered as much as collecting and coding stages TUS are a rich and vast source of data But underexploited in general While they are costly
Similar presentations
© 2025 SlidePlayer.com. Inc.
All rights reserved.