Association Cross Tabulation (Cross tabs)
There are two major types of data based on the nature of the data. One is categorical the
other is interval.
GMAT scores are interval data. This means that GMAT scores are sequential numbers,
and the magnitude of the value between the numbers is always equal to a constant,
usually 1. The data can be added and subtracted and means can be calculated. The
average GMAT score for BYU is about 630. The average GMAT for Stanford is about
740.
Gender is categorical data; individuals are almost always either male or female.
The count of male and female can be accomplished but you can not do much else with
the data other than counting the number in each group.
When a marketing researcher is analyzing categorical data he can not use correlation to
measure the association.
With categorical data you use cross tabs (cross tabulations). The researcher may be
interested in something like the likelihood of purchase of individuals based on gender.
Are males or females more likely to buy the product?
If the researcher has asked the questions:
How likely you are to buy new product X:
____Highly likely
____Somewhat likely
____Not Likely
____Highly unlikely
What is your gender?
____ Male ____Female
The researcher will end up with everyone falling into one of eight classifications based on
the responses to the above two questions.
HL SWL SWUL HUL
FEMALE
MALE
If a count was made for the 101 respondents the table would look like:
HL SWL SWUL HUL TOT
M 5 18 14 17 54
F 10 18 18 1 47
The above is an example of cross tabulation
It can be created by counting but a more efficient process is to use a statistical software
program. Excel has pivot tables. Pivot Tables are hard to work with.
In industry most marketing researchers use SPSS. It is in the computer lab at the Marriott
School. It can also be down loaded in a limited version from the SPSS web site.
To do a cross tab in SPSS go to the tutorial called SPSS TOOL KIT. This tutorial shows
how to create a data file and do a cross tab analysis.