You've probably seen the data. You get a spreadsheet
that has survey data with a row for each survey
returned and a column for each question on the
survey. The answers to some of the questions
are in the form of "Excellent", "Good",
"Fair" and "Poor" while
others are "Yes", "No" or
"Not Applicable". And to make things
even more complicated, in each column there
could be a number of blank spaces…How can
I produce a chart from these Data? HELP!
OK. How will we get measurable information
from this survey with text as our data? First,
we ask, "What is the average response?"
What you need to do is translate those answers
(Excellent, Good, Fair, Poor) to numbers. For
the Yes/No answers you may want to find the
percent that answered "Yes" where
the question was answered and applicable. Therefore,
you would add up all the Yes answers and divide
it by the number of Yes and No answers. Again….HELP!
It's not unusual to see data stored in this
manner, but it makes it difficult for the responsible
party to analyze. You could do tricks in Excel
to help you in the analysis, but that can be
tedious and time consuming, more than likely
adding columns to your spreadsheet that are
unnecessary. If you use a database as a data
source, you may be further restricted in what
you can and cannot do…Going down for the
last time…HELP!
Statit Performance Indicator
& Management Dashboard (piMD) is here
to HELP you! With our "Recode Feature",
it is a simple matter of telling Statit piMD,
for example, that any time an "Excellent"
occurs in this column, interpret it as a 4,
a "Good" as a 3, etc. Perhaps we want
to count the number of "Excellent"
answers for all surveys where that question
was answered, so we would recode the "Excellent"
responses as "1", "Good",
"Fair" and "Poor" as zero
(0) and anything else as blank. Finally, when
you see a "Yes" in a column, interpret
that as a "1" and a "No"
as "0" while anything else would go
unused in our analysis.
One of our Statit customers has a set of survey
data with almost 3400 surveys with 32 questions
where the possible answers are "Yes",
"No", "Not Applicable" or
blank. Can you imagine setting up this spreadsheet
to transform that data on an ongoing basis?!
Doctor, please prescribe two aspirin and time-off
in Tahiti!!!
No one to fill the prescription? Don't worry…simply
using the Recode feature in Statit piMD, our
customer transformed those 3400 surveys and
32 questions in less than 5 minutes. And this
change is now persistent and would be maintained
for additional surveys added to the spreadsheet
with no other intervention! Is this even better
than the time off in Tahiti? OK…just wanted
to make sure you were still reading!!
The Recode feature lets the responsible party
decide how they want to analyze the data. Do
you want the percent of "Yes" answers
for all surveys returned or all surveys returned
with either a "Yes" or a "No"
answer to a particular question?
It's also not unusual to see something like
M for Male and F for Female in database tables.
Again, it is extremely simple to recode those
character-based attributes to a numeric for
easy counting and summing.
Here is an example of the way the data is stored
in a survey spreadsheet:
This is what it could look like using the Recode
feature in Statit piMD:
This form is much better for numerical analysis.
The Yes response now shows up as 1, the No response
shows up as 0, and all others are blanks and
will (typically) not be used in calculations.
Sums, averages and other arithmetic functions
are much easier to accomplish. And it requires
no change from the way the responsible party
is collecting data.
That leads us to easily create a chart such
as:
Often times, survey data is retained as text
or characters. Measuring and analyzing data
in this matter is difficult if not impossible.
You are already too busy to have to deal with
this in Excel or some other way of transforming
the data. "Outsource" this to Statit
piMD and think about that trip to Tahiti…with
Statit piMD, you'll have time to get away!
For more information about Statit piMD and
the Recode feature, please call 541.752.4100
or send email to
.
