Comments on: Validity in Research Design

By: Joe Maxwell

Joe Maxwell — Thu, 29 Dec 2011 01:26:26 +0000

Tariq’s post is another attempt to force qualitative research into a quantitative framework. However, rejecting his argument doesn’t require abandoning the concept of validity in qualitative research; it just means that validity needs to be defined in ways that fit what qualitative researchers actually do. See the entry on Validity in the Sage Encyclopedia of Qualitative Research (Sage Publications, 2008). I think most qualitative researchers accept that some conclusions are more credible, trustworthy, and useful than others, and there are a number of different approaches to assessing this that are currently in use.
A currently prominent view is that validity is a property of conclusions, not methods or designs, because any method or design can result in valid conclusions in some situations and invalid ones in others. Methods are obviously important, but they don’t assure the validity of the results; the latter requires testing your conclusions against plausible alternative conclusions. I can’t give the full argument for this here; see the chapter on Understanding, Validity, and Evidence, in my book A Realist Approach for Qualitative Research (Sage, 2011).

By: Drwho314159265358979

Drwho314159265358979 — Thu, 01 Dec 2011 16:57:00 +0000

Tanya Porter,

You’re partially right…qualitative research is not concerned with reliability. However, many qualitative researchers reject the idea of validity being a requirement of qualitative inquiry, because they reject the assumption that there is an objective reality outside of our perception, and therefore, it makes no sense to be concerned with the “truth”
or “falsehood” of an observation.

Guba and Lincoln gave a set of four alternatives to validity, which can be used by qualitative researchers. They are summarized on this page:

http://www.socialresearchmethods.net/kb/qualval.php

By: Tonya Porter

Tonya Porter — Thu, 01 Dec 2011 16:12:00 +0000

Qualitative disregards RELIABILITY not validity. It does not have to replicable because behaviour is often not replicable but it is valid to what the researcher has set out to measure and observe.

By: atiye j

atiye j — Fri, 22 Jul 2011 17:43:00 +0000

i am writing a dissertation using stickers, how can i pre-test and validate it.

By: Drwho314159265358979

Drwho314159265358979 — Tue, 19 Jul 2011 22:32:00 +0000

“…some qualitative researchers have gone so far as to suggest that
validity does not apply to their research even as they acknowledge the
need for some qualifying checks or measures in their work. This is
wrong. To disregard validity is to put the trustworthiness of your
work in question and to call into question others confidence in its
results. Even when qualitative measures are used in research they need
to be looked at using measures of reliability and validity in order to
sustain the trustworthiness of the results.”

By supplying this (reasonably accurate, if incomplete) list of different kinds of validity, you have not supported your argument that qualitative research must be shown to be both valid and reliable.

“Any research worth its weight is concerned with whether what is being measured…”"

It is precisely because there are no measurements made in qualitative
research that validity and reliability are not (in fact, cannot be) calculated or tested.

By: Drwho314159265358979

Drwho314159265358979 — Tue, 19 Jul 2011 22:18:00 +0000

Two things:

1. I would suggest placing your definitions within some organizational structure to add clarity (check Wikipedia’s article on validity). For example, you could distinguish between validity of measurement (construct, content and criterion), and validity of experimental design (all the others, plus construct validity); also, you could subordinate them the way you did with criterion-related validity and its two types (so that face validity is shown to be a type of content validity, for example).

2. Your definition of concurrent validity conflates the two subtypes of concurrent validity (discriminant and convergent); where discriminant validity compares scores for groups that do and do not have a characteristic, and convergent validity refers to the extent of correlation of related variables.

By: Irfan Malik

Irfan Malik — Wed, 20 Apr 2011 06:41:00 +0000

this is the total theory about the validity i am interesting to know that what statistical tools are applicable and how they can be applied? and one thing that is missing in this article is lack of examples if you include the practical example with the theory it will make it very usefulnregards nIrfan MaliknPakistan

By: Mus

Mus — Thu, 17 Feb 2011 12:28:58 +0000

Please I’m writing a thesis and want to develop a questionnaire form for a survey I want to conduct. In constructing my the questionnaire what should I include or avoid to ensure the questionnaire stands the test of validity?. The topic of the thesis borders on Challenges of internet use in Africa.

By: Obert

Obert — Sat, 21 Aug 2010 14:57:55 +0000

Does the concept of internal validity (history, maturation, instrumentaility etc) apply in qualitative business research? If they dont what are the threats to internal validity in qualitative business research?

By: Habib

Habib — Sun, 18 Apr 2010 16:27:01 +0000

Construct validity refers to validity of the ideas and concepts used in the research. Constructs are “concepts” and “ideas”. This particular validity is often problematic and not clear cut. There is a subjective (interpretive) element in it that makes it somewhat debatable.