Qualitative Data Uses
Qualitative descriptions help you record changes within your experiment
that may not necessarily be measureable. When collecting these observations,
you describe how something looks, smells, feels, sounds, or tastes (when
appropriate) or categorize it into a specific category. However, just because
you may not be using numbers, don’t lose your objectivity. Your observations
should be scientific in nature and not make judgments or inferences.
An inference is a conclusion, based on facts, that is perceived to be true by the
researcher. Be careful, however, when you make an inference. The statement
“The solution looks normal” is an inference, but this conclusion obviously is
based on observations that are not recorded. Although you may know what
you mean, inferences written without factual descriptions will not help you
compare results at the end of the experiment. Statements that include inferences
are best saved for after data are collected. Instead, the actual observations,
which lead to the inferences, should be recorded. Remain scientific and
use detailed and descriptive language.
If you have trouble determining how to describe qualitative data, ask
yourself, “What is ‘normal’ about this?” Make a long list of adjectives to
describe the qualitative aspects of your dependent variable. For example, if
you are studying viscosity of a fluid, list words that will help you describe
varying thicknesses of the solutions—for example, stringy, thready, dense,
clumps, or runny. If you photograph the entities throughout the experiment,
you’ll be able to compare qualitative differences. You may notice something
in photographs that you didn’t notice on a day-to-day basis. These observations
will help supplement the quantitative data that you collect.
In addition to narrative descriptions, qualitative data can also be in the
form of category frequency or ratings, both of which use numbers. Counting
frequencies allows you to keep track of changes that are not normally quantified.
For example, to record color change, you could use paint swatches, with
each gradient of color assigned consecutive numbers—perhaps low numbers
for lighter shades and higher numbers for darker shades. In a catapult-testing
experiment, after research and/or pretrials, you might determine that there
are three basic arch shapes in which the projectile might fall. After each trial,
you could measure distance (quantitative) but also determine which of the
three arch categories a catapult belongs to (qualitative).
If you choose to do behavioral research, you might collect data on location,
like at a zoo, for animal behavior or at a coffee shop for human behavior.
Recording behavior is a good time to use qualitative data. Behavioral research
can be recorded several ways; the most common are focal sampling, scan
sampling, and sequence sampling (Morgan 2009).
• In focal sampling, you choose one individual or group of individuals
and record your observations for a set length of time. You watch and
record everything you observe, writing in a narrative form.
• In scan sampling, you record the activity of an individual or group
at preselected time intervals. Scan sampling should give you a
sample representation of the behaviors taking place, and if you
predetermine categories, it will also allow you to tally behaviors that
can be used in data analysis. For example, if using scan sampling in
the river otter experiment (Figure 2.1), you might observe the river
otter in two-minute segments for several hours. At the moment
each two minutes has passed, you would record what the otter
is doing. Otter behavioral categories to be tallied might include
walking, swimming underwater, floating on back, diving, grooming,
foraging, or playing. Scan sampling helps keep an accurate record
of observed behaviors as well as a record of changes over time if
multiple observations are made.
• In sequence sampling, you record behaviors that occur within a
sequence, in the order in which they occur. The rubric in Table 2.1
is an example of sequence sampling. The rubric was designed for a
horse-training experiment in which the researcher wanted to keep
track of a horse’s progress as it learned a new skill. The behavior
(taking a first step) was broken down into smaller pieces and then
used during each training session to record the progress of the horse
as it learned the new behavior.