Dear Don,
You wrote:
-snip-
In the form of Human-Centered Design that i practice and preach, which I
now call Observe, Make, Test (TOM -- letters restructured to make it
pronounceable):
- Observations are qualitative
- Making can be sketching, drawing, prototyping -- think of this as
instantiation
- Testing transforms the qualitative Observations into quantitative
information, via the Made material
-end snip-
The sequential mixed methods research design you propose seems to me to be
quite similar to Blessing and Chakrabarti’s (2009) DRM: Design Research
Methodology. Blessing and Chakrabarti’s (2009, p. 15) methodology has four
stages:
1. Research Clarification through literature analysis to identify goals
2. Descriptive Study 1 though empirical data analysis to understand the
factors structuring/generating something in the world
3. Prescriptive Study through synthesis (designing) to create support (a
design)
4. Descriptive Study 2 though empirical data analysis to investigate the
impact of the support and its ability to realise the change it is intended
to make.
As I see it, the difference between DRM and TOM is that Blessing and
Chakrabarti do not specify that first descriptive study should collect
qualitative data and the second descriptive study should collect
quantitative data, they leave it up to the researcher to choose the best
research method for the particular situation being investigated. I quite
like Blessing and Chakrabarti’s methodology and have used it in teaching
graduate level strategic design. However, we usually do not have time to
get through all four stages in a single semester so we tend to focus on
stages 1-3, and students often choose to collect either qualitative or
quantitative data depending on which methods they are most familiar with
using.
Qualitative data is useful for exploring new information and quantitative
data is useful for resolving the problem of sampling error through choosing
large enough samples and choosing them wisely. However, both approaches
also have potential pitfalls. Basing design claims on qualitative data may
not address the actual needs of users and quantitative research cannot be
performed where there is a lack of specific fundamental knowledge relevant
to the area of design to base hypotheses on.
In my undergraduate level introduction to design research course I teach an
intentionally mixed methods research design with stages as follows:
1. Research Clarification through literature analysis to identify goals
2. Descriptive Study 1 though qualitative empirical data analysis to
understand the something in the world
3. Descriptive Study 2 though quantitative empirical data analysis to test
the generalizability of the insights from Descriptive Study 1
4. Prescriptive Study through synthesis (making) to create support (a
product)
In this research design there is a transformation step where the
qualitative insights are transformed into hypotheses that can be tested
using quantitative instruments. Because the insights from DS1 are
transformed into hypotheses that are tested in DS2, I call this research
design an integrated sequential exploratory mixed methods approach. The
intention is that the qualitative analysis discloses insights relevant to
specific contexts and quantitative analysis makes it possible to test the
insights to derive some understanding of their certainty.
There are many ways to use qualitative and quantitative data in research.
Mixed method research refers to the use of both qualitative data and
quantitative data within one research design to answer the same research
question. Different approaches within mixed method research can be
distinguished based on either priority or implementation of data
collection. In essence, priority can be given to either qualitative or
quantitative research or equal weight can be placed on both within the
research design. Implementation of data collection refers to the choice to
either collect or interpret qualitative and quantitative data concurrently
or sequentially. The major mixed method designs derived from combinations
of these two factors are: triangulation, embedded, explanatory and
exploratory.
Triangulation occurs when both qualitative data and quantitative data are
interpreted simultaneously to provide more reliable results. Embedded
research seeks to clarify the results obtained with one type of research
with the other type of research. This can happen either sequentially or
concurrently and the choice of which one is used to clarify the other
depends on the research question. In an explanatory research design a
quantitative research phase is followed by a qualitative phase whereby the
qualitative results explain the quantitative results. The quantitative
phase informs the questions or sampling of the qualitative phase.
Exploratory designs start with qualitative research and those findings are
subsequently validated by quantitative results. Typically, the factors or
outcomes identified in the qualitative phase are applied to a larger and
more diverse sample in the quantitative phase. This latter approach is
often employed in relatively unstudied areas.
I think that the nature and uniqueness of situations often addressed by
designers call for an exploratory research approach. The qualitative
research can identify factors relevant for the specific context, while the
quantitative research can then validate the certainty of those factors and
test the design claim put forward. When qualitative and quantitative
research is integrated within a research design, the claims produced are
both valid and reliable.
I think it would be interesting to see what kind of outcomes you would get
if you used the DRM approach but with mixed methods for both DS1 and DS2.
So that would be the same as your TOM research design but using mixed
methods for both the O and the T stages.
Best regards
Luke
Blessing, L. T. M., & Chakrabarti, A. (2009). DRM, a design research
methodology. Dordrecht; London: Springer.
On 18 May 2014 01:24, Don Norman <[log in to unmask]> wrote:
> Roger Martin of U. Toronto and I just had an on-stage discussion at the
> IIT/Institute of Design Strategy Conference in Chicago (moderated by
> Patrick Whitney).
>
> Roger made a very interesting point about the need to combine both
> qualitative and quantitative information (he called these "intuitive" and
> "analytical" -- but I detest the word "intuitive" because it doesn't mean
> what most people think it means)
>
> During the discussion of this, a new insight (at least for me) emerged:
> that the transformation between qual and quant was via testing.
>
> In the form of Human-Centered Design that i practice and preach, which I
> now call Observe, Make, Test (TOM -- letters restructured to make it
> pronounceable):
>
> - Observations are qualitative
> - Making can be sketching, drawing, prototyping -- think of this as
> instantiation
> - Testing transforms the qualitative Observations into quantitative
> information, via the Made material
>
> This argument requires considerable elaboration, but I wondered if this
> gorup can provide constructive critique of the notion.
>
> Part of this is to try to transform the argument about quantitative versus
> qualitative to eliminate the word "versus" with something else ("combined
> with"?). The point is that each serves a different purpose, and both are
> often needed.
>
> don
>
--
Luke Feast | Lecturer | Early Career Development Fellow | PhD Candidate |
Faculty of Health, Arts and Design, Swinburne University of Technology,
Melbourne, Australia | [log in to unmask] | Ph: +61 3 9214 6165 |
http://www.swinburne.edu.au/health-arts-design/
-----------------------------------------------------------------
PhD-Design mailing list <[log in to unmask]>
Discussion of PhD studies and related research in Design
Subscribe or Unsubscribe at https://www.jiscmail.ac.uk/phd-design
-----------------------------------------------------------------
|