Session: Estimating and correcting for measurement error

Time: Wednesday 4th July, 11:15 - 12:45

Convenor:

Dr Alexandru Cernat (Unviersity of Manchester)

Abstract Details

While the study of the social sciences has developed important concepts such as social class, attitudes towards politics or globalisation have been proposed. Unfortunately, often such concepts are not measured directly but through proxies such as survey questions. The difference between our concepts of interest and the observed/collected data is called measurement error. It has been shown that measurement error can lead biased results in statistical analysis if ignored. In the session we will discuss about what is measurement error, how to correct for it and what new methods are being developed in this field.

Presentation downloads

Presenter: Melanie Revilla

Correction for measurement errors: how to do it and what is the impact?

The level of the session is: Accessible

Presentation details

Presentation 1

Start time: 11:15

Presentation title: What is measurement error and how can we estimate it?

Presenter:

Dr Daniel Oberski (University of Utrecht)

Presentation 2

Start time: 11:35

Presentation title: Correction for measurement errors: how to do it and what is the impact?

Presenter:

Dr Melanie Revilla (Universitat Pompeu Fabra)

Presentation 3

Start time: 11:55

Presentation title: Multilevel structural equation modelling of longitudinal MTMM data

Presenter:

Dr Tobias Koch (Leuphana University of Lüneburg )

Presentation 4

Start time: 12:20

Presentation title: Estimating multiple types of measurement errors concurrently using the multitrait multi error approach

Presenters:

Dr Alexandru Cernat (University of Manchester)
Dr Daniel Oberski (University of Utrecth)