Correlates of Consent

Loading...

Del curso dictado por University of Maryland, College Park

Combining and Analyzing Complex Data

31 calificaciones

University of Maryland, College Park

Combining and Analyzing Complex Data

31 calificaciones

Curso 6 de 7 en SpecializationSurvey Data Collection and Analytics

In this course you will learn how to use survey weights to estimate descriptive statistics, like means and totals, and more complicated quantities like model parameters for linear and logistic regressions. Software capabilities will be covered with R® receiving particular emphasis. The course will also cover the basics of record linkage and statistical matching—both of which are becoming more important as ways of combining data from different sources. Combining of datasets raises ethical issues which the course reviews. Informed consent may have to be obtained from persons to allow their data to be linked. You will learn about differences in the legal requirements in different countries.

De la lección

Ethics

This module will discuss key issues in obtaining consent to record linkage. Failure to consent can lead to bias estimates. Current research examples will be given as well as practical suggestions on how to obtain linkage consent.

Linkage Consent and Consent Bias8:02

Correlates of Consent3:37

Bias in Administrative Estimates2:07

Optimizing Linkage Consent6:24

Conoce a los instructores

Richard Valliant, Ph.D.
Research Professor
Joint Program in Survey Methodology

So, Correlates of Consent.

We talked about that some people do give consent and others don't.

And it would be interesting, and therefore, a lot of papers,

you see them listed on the bottom of the slides have looked into this.

What correlates with people giving consent?

Are there any particular characteristics of the respondents,

certain demographic groups, certain income groups that could be a problem here?

And likewise, are there interviewer characteristics?

And indeed there are.

We have seen that unexperienced interviewers tend to

have lower consent rates.

If the interviewer herself is willing to share data and

to have her own data linked, there is a higher consent rate among

those people interviewed by that particular interviewer.

It's very interesting if you think about that consequence, right?

So if you have a study that focuses on

that kind of linkage and having different data sets together.

You should want to make sure that the interviewer is trained and

really, fully embraces that concept.

Now the list is long of potential respondent characteristics.

Here you can see the direction of these variables on the consent rates.

So in the study from Nancy Bates, from 2005,

she saw that there was a lower consent rate among

females, a negative correlation with age,

negative correlation with education and earnings.

Very different though, some studies done a little earlier by Jenkins and

Young you see opposites effect of education.

So bottom line, looking at this picture, some studies show correlates,

others don't.

And I guess with the example of net worth, you find studies that show

it in all directions, which is leaving us a little in the dust of what happens.

And this could be in part because it depends on what kind of survey

we are talking about.

Now here's an interesting analysis that Daniel Yang, and Scott Fricker,

and John Eltinge did with data from the Bureau of Labor Statistics.

Particularly, they used the consumer expenditure as part of

the interview survey, to look at biases in the survey estimates.

Here, they had a consent rate of about 80%.

What you see in these two columns, respondent mean and

consenting or the mean among the consenting units is for five variables,

family income, vehicle cost, property taxes, property value, rental value.

How the average changes among the consenters.

So on family income, It

appears that people with lower income consented less.

And therefore we have, if we only analyze the consenting units,

a higher respondent mean.

That's the way you interpret this table.

So differences are not huge, but some of these differences are indeed significant.

And that is the third column here, so

the bold faced values in that third column are the significant differences.

Explora nuestro catálogo

Inscríbete de manera gratuita y obtén recomendaciones personalizadas, actualizaciones y ofertas.

Coursera brinda acceso universal a la mejor educación del mundo, al asociarse con las mejores universidades y organizaciones, para ofrecer cursos en línea.

© 2019 Coursera Inc. Todos los derechos reservados.

Descargar en la App Store

Consíguelo en Google Play