|
research
design
Research findings must focus on how they impact your
objectives , they must be solution
oriented, and they
must be actionable. They must tell
you what action
you need to take to meet your objectives. Therefore,
we often work backwards from the decisions that need to be made in designing research
methodologies. Below are issues we address:
What
are the questions that need to be asked?
What is the appropriate way to ask them so that we get answers that are
meaningful?
What is
the best source to get the answers?
What are the best research methodologies to give us actionable information?
What are the best methods to analyze the data collected to derive meaning?
How do we
get from meaning to actionable information?
We spend a considerable
amount of time with each client discussing what they want to achieve and how
they see their business climate. We look for hidden assumptions that need
to be tested to insure the answers we provide will help our client make
better decisions. Before we design a survey, we sometimes conduct some form
of qualitative research, such as in person interviews or focus groups, with
existing or potential customers/constituents to be sure our client’s
perceptions of issues are in sync with their market/constituency.
Return to top
The way questions are
phrased can greatly influence the kinds of responses we get. In one study,
performed by the Institute of Survey Research at the University of Michigan,
researchers found a 15% difference in how people responded to a question
simply based on changing the word “prohibited” to “banned” in an item.
We must even consider
whether respondents are likely to know the answers. Based on our own
experience, we know that people will often answer questions even when they
no little or nothing about the topic.
We have over twenty-five
years of experience designing questions that will provide useful, unbiased
answers. We also pretest questionnaires to be sure that they are easily
understood by respondents.
Return to top
Sometimes getting useful answers requires that we look at existing records,
sometimes that we talk to people, and sometimes that we observe
peoples' behavior. We are innovative and creative in looking at
alternatives. On one project we videotaped license plates of vehicles
exiting mall parking lots to identify a large group of mall patrons and
their method of access. On another project, we brought our expertise to the
analysis of “slamming” records to show the Texas Public Utility commission
that a long distance carrier’s performance in minimizing “slamming”
incidents was exemplary and not poor, as originally perceived.
Return to top
We match what our client
needs to know with the best way to obtain the information. We either
use qualitative methods, such as executive interviews, or quantitative
methods, such as surveys, or a combination of these.
Return to top
The point of data
analysis is to discover patterns in the data that provide the answers we
need. Often the simplest statistical tools are best able to do this. On
other occasions more sophisticated methods are required. Our philosophy is
to use the simplest method that will give the answers we need. The reason
for this is that the tools are often more intuitive for the client. We have
found that the easier the method is for the client to understand the more
likely it is that the results will be used effectively.
Having said that we do
use a variety of data analytic methods including simple cross tabs,
univariate and multivariate analysis of variance, multiple regression
analysis, factor analysis, multidimensional scaling and conjoint analysis.
Return to top
For research to be of any value, it must give us the insight to make good
decisions. We might find meaning but not know how to use it to take
productive action. Almost twenty years of experience across many
industries has given us the insight to help our clients take the final step
from understanding information we give them to understanding what actions
will move them toward their objectives.
Return to top
|