Microsoft Office Tutorials and References
In Depth Information
we offered some general advice on the collection and presentation of quantitative
data. It is worth repeating that advice at this point, but now from the perspective of
qualitative data presentation.
1. Not all data are created equal —Spend some time and effort considering the type
of data that you will collect and how you will use it. Do you have a choice in
the type of data? For example, it may be possible to collect ratio data relating
to individual’s annual income ($63,548), but it may be easier and more conve-
nient to collect the annual income as categorical data (in the category $50,000 to
$75,000). Thus, it is important to know prior to collection, how we will use the
data for analysis and presentation.
2. More is better —If you are uncertain of the speciﬁc dimensions of the obser-
vation that you will require for analysis, err on the side of recording a greater
number of dimensions. For example, if an invoice in our payment data (see
Table 4.1) also has an individual responsible for the transaction’s origination,
then it might be advisable to also include this data as a ﬁeld for each observa-
tion. Additionally, we need to consider the granularity of the categorical data
that is collected. For example, in the collection of annual income data from
above, it may be wise to make the categories narrower rather than broader: cate-
gories of $50,000–$75,000 and $75,001–$100,000 rather than a single category
of $50,000–$100,000. Combining more granular categories later is much easier
than returning to the original source data to collect data in narrower categories.
3. More is not better —If you can communicate what you need to communicate with
less data, by all means do so. Bloated databases and presentations can lead to
misunderstanding and distraction. The ease of data collection may be important
here. It may be much easier to obtain information about an individual’s income
if we provide categories, rather than asking them for an exact number that they
may not remember or want to share.
4. Keep it simple and columnar — Select a simple, descriptive, and unique title
for each data dimension (e.g. Revenue , Branch Ofﬁce , etc.), and enter the data
in a column, with each row representing a record or observation of recorded
data. Different variables of the data should be placed in different columns. Each
variable in an observation will be referred to as a ﬁeld or dimension of the obser-
vation. Thus, rows represent records and columns represent ﬁelds. See Table 4.1
for an example of columnar formatted data entry.
5. Comments are useful —It may be wise to include a miscellaneous dimension
reserved for general comments—a comment ﬁeld. Be careful, because of the
variable nature of comments, they are often difﬁcult, if not impossible, to query.
If a comment ﬁeld contains a relatively limited variety of entries, then, it may not
be a general comment ﬁeld. In the case of our payment data, the comment ﬁeld
provides further speciﬁcity to the account information. It identiﬁes the project
or activity that led to the invoice. For example, we can see in Table 4.1 that the
record for Item 1 was Ofﬁce Supply for Project X . Since there is a limited number
of these project categories, we might consider using this ﬁeld differently. The
title Project might be an appropriate ﬁeld to record for each observation. The
Comment ﬁeld could then be preserved for more free form data entry.