Microsoft Office Tutorials and References
In Depth Information
In Excel, the internal area of the cross-tabulation table is referred to as the data
area , and the data elements captured within the data area represent a count of
respondents (7, 7, 4, 11). The dimensions are referred to as the row and column .
On the margins of the table we can also see the totals for the various values of each
dimension. For example, the data area contains 11 total bad respondent preferences
and 18 good , regardless of the gender of the respondent. Also, there are 14 total
females and 15 males, regardless of their preferences.
The marginal dimensions are selected by the user, and the sub-totals, for example
11 Bad opinions, can be useful in analysis. In the data area we currently display a
count of respondents, but there are other values we could include, for example, the
respondents average age or the total income for respondents meeting the row and
column criteria. There are also many other values that could be selected. We will
provide more detail on this topic later.
We can expand the number of data elements along one of the dimensions, row
or column , to provide a more detailed and complex view of the data. Previously, we
had only Gender on the row dimension. Consider a new combination of Region and
Gender . Region has 4 associated categories and Gender has 2; thus, we will have
8(4
2) rows of data, plus a totals row. Table 5.4 shows the expanded and more
detailed cross-tabulation of data. This new breakdown provides detail for Male and
Female by region. For example, there are 3 females and 6 males in the East region.
There is no reason why we could not continue to add other dimensions, either to
the row or column , but from a practical point of view, adding more dimensions
can lead to visual overload of information. Therefore, we are careful to consider
the confusion that might result from the indiscriminant addition to dimensions. In
general, two characteristics on a row or column are usually a maximum for easily
understood presentations.
You can see that adding dimensions to either the row or column results in a data
reorganization and different presentation of Table 5.2; that is, rather than organizing
based on all respondents (observations), we organize based on the speciﬁc cate-
gories of the dimensions that are selected. All our original data remains intact. If
we add all the counts found in the totals column for females in Table 5.4, we still
×
Table 5.4 Cross-tabulation
of gender/region and product
1 preference in terms of
respondent count
Product 1
Region
Gender
Good
Totals
East
Female
1
2
3
Male
2
4
6
West
Female
2
2
4
Male
0
2
2
North
Female
1
1
2
Male
1
4
5
South
Female
3
2
5
Male
1
1
2
Totals
11
18
29

Search JabSto ::

Custom Search