Decoding Variable Types in Data Lesson
Student Handout
INTRODUCTION
Collecting data is essential for answering research questions. When designing a study, researchers must decide
what types of variables to collect data on and how to measure them. However, direct measurements (“raw
data”) are rarely in the appropriate form for answering detailed questions. Researchers often need to perform
calculations or convert data in preparation for analysis.
In this activity, you will examine decisions about variable types and data conversions using a dataset from a
research study on how lizards responded to a high-elevation (low-oxygen) environment. You will explore
different ways of using these data to answer research questions.
• In Parts 1 and 2, you will use the lizard dataset to distinguish between types of numerical and categorical
variables. You will also use the data to perform and evaluate data conversions.
• In Part 3, you will design a study to answer your own research question. This will include proposing which
variables to collect data on.
MATERIALS
• calculator or access to a spreadsheet program
BACKGROUND
The dataset in this activity was collected from an experiment with the common wall lizard (Podarcis muralis).
This species is native to southern Europe, and it is found in rocky habitats across a broad geographic range. The
lizards usually live at lower elevations, where temperatures are warm enough for survival and reproduction. But
as temperatures increase due to global climate change, the lizards could move into higher-elevation
environments that were formerly too cold — if they can handle other high-elevation conditions, such as lower
oxygen.
To investigate how the health and physical performance (physiology) of the lizards respond to higher elevations,
researchers conducted the experiment illustrated in Figure 1. First, they collected adult lizards at six low-
elevation (425–522 m) locations around France. They took about half of the lizards to a research lab at a low
elevation (436 m). The other half were taken to a lab at a higher elevation (2,877 m), which had 25% less
oxygen. The scientists then collected measurements, such as running endurance, from the lizards at both labs.
You’ll learn more about these measurements in Parts 1 and 2 of the activity.
Figure 1. An illustration of the experimental design for the lizard study (Gangloff et al. 2019).
[Link] Published November 2024
Page 1 of 7
Lesson
Decoding Variable Types in Data Student Handout
PART 1: The Lizard Dataset and Body Mass
Researchers measured characteristics of the lizards at both labs at different timepoints. Some of their data are
shown in Table 1.
Table 1. A subset of data from Gangloff et al. (2019). The lizards in this subset were collected from two low-
elevation sites in France (Alas and Engomer) and tested at either the low- or high-elevation lab. Body mass and
endurance for each lizard were measured at three timepoints (T1 = first day of the experiment, T2 = 3 weeks
after T1, T3 = six weeks after T1).
Column A Column B Column C Column D Column E Column F
1 Lizard ID Collection site Lab elevation Timepoint Body mass (g) Endurance (s)
2 1123 Alas High T1 7.18 84.80
3 1123 Alas High T2 7.43 137.84
4 1123 Alas High T3 7.26 149.08
5 232 Alas High T1 5.01 83.76
6 232 Alas High T2 4.92 87.76
7 232 Alas High T3 4.81 56.12
8 246 Alas Low T1 7.54 60.04
9 246 Alas Low T2 7.79 122.12
10 246 Alas Low T3 8.16 162.88
11 245 Alas Low T1 5.52 60.04
12 245 Alas Low T2 6.37 144.00
13 245 Alas Low T3 6.01 149.08
14 228 Engomer High T1 4.59 49.48
15 228 Engomer High T2 4.88 259.40
16 228 Engomer High T3 4.68 172.76
17 225 Engomer High T1 6.75 70.20
18 225 Engomer High T2 7.15 139.68
19 225 Engomer High T3 6.93 100.92
20 255 Engomer Low T1 5.5 115.88
21 255 Engomer Low T2 6.48 195.12
22 255 Engomer Low T3 6.05 214.40
23 252 Engomer Low T1 6.69 113.32
24 252 Engomer Low T2 6.84 112.52
25 252 Engomer Low T3 6.93 153.72
[Link] Published November 2024
Page 2 of 7
Lesson
Decoding Variable Types in Data Student Handout
1. Identify the following variables from Table 1 as either categorical or numerical. Further classify each
variable’s specific type as nominal, ordinal, binary, discrete, or continuous.
Variable Categorical or numerical? Specific type (nominal, ordinal, binary,
discrete, or continuous)
Collection site categorical nominal
Lab elevation categorical binary
Timepoint categorical ordinal
Body mass (g) numerical continuous
Endurance (s) numerical continuous
Let’s focus on the variable of body mass, which is measured by weighing the lizards. When lizards are moved to
high elevations, which have less oxygen, their body mass may change due to changes in their muscle mass and
fat stores. This can reflect different metabolic costs of living in low- vs. high-elevation environments. (All the
lizards in the study had an unlimited supply of the same food.)
2. Consider this research question: “Does lizard body mass differ over time between the low- and high-
elevation labs?”
a. Which variables in the dataset (Table 1) would you use to answer this question?
body mass, timepoint, lab elevation and lizard id
b. Generate a prediction for the result you expect.
lizards at high-elevation labs will lose some body mass after
some time because at high levels there is less oxygen, while at low-elevation lab they will keep or gain their mass
Researchers must decide how to analyze their measured variables to answer their questions. Often, there are
multiple options. For example, here are two ways to use the data in Table 1 to analyze how elevation affects
lizard body mass:
● Body mass at one timepoint: Compare body mass (Column E) at timepoint T2 or T3 for lizards at low vs. high
lab elevation.
● Change in body mass: Calculate the change in body mass for each lizard between two timepoints (for
example, body mass at T3 − body mass at T1), then compare the results for lizards at low vs. high lab
elevation. This is an example of a data conversion.
3. Why might “change in body mass” be more useful for answering the research question (“Does body mass
differ over time between lizards at the low- and high-elevation labs?”) than “body mass at one timepoint”?
because it shows how each lizard’s mass changes after some time. this does not keep in
account the differences between the lizards at the start so the comparison is more fair
There are also multiple options for calculating a change in body mass. Let’s explore two different methods, using
Lizard 1123 (Rows 2–4 in Table 1) as an example.
[Link] Published November 2024
Page 3 of 7
Lesson
Decoding Variable Types in Data Student Handout
Method 1: Change in Body Mass in Grams (g)
General equation:
change in grams = body mass at T3 − body mass at T1
For Lizard 1123:
change in grams = 7.26 g − 7.18 g = 0.08 g
Method 2: Percent Change in Body Mass
General equation:
body mass at T3 − body mass at T1
percent change = × 100
body mass at T1
For Lizard 1123:
7.26 g − 7.18 g
percent change = × 100 = 1.1%
7.18 g
Thus, Lizard 1123 gained 0.08 g from T1 to T3, which is a 1.1% gain in body mass.
4. Perform the change in body mass calculations for the three other lizards in Table 2. Record your answers in
Columns E and F of the table.
Table 2. Data for four lizards. Columns A–D are from the data in Table 1. (As In Table 1, T1 = the first day of
the experiment and T3 = six weeks after T1.) Column E shows “change in body mass in grams (g)” calculated
using Method 1. Column F shows “percent change in body mass” calculated using Method 2.
Column A Column B Column C Column D Column E Column F
Lab Body mass at Body mass at Change in body Percent change
Lizard ID
1 elevation T1 (g) T3 (g) mass (g) in body mass (%)
2 1123 High 7.18 7.26 0.08 1.1
3 232 High 5.01 4.81 -0.2 -4.0
4 246 Low 7.79 8.16 0.37 4.8
5 245 Low 5.52 6.01 0.49 8.9
5. What does a negative value in Column E or F indicate?
it means that the lizard lost weight
6. Which do you find more useful for describing change in body mass, change in grams (Column E) or percent
change (Column F), and why?
percent change because like that the lizards can be of different sizes and the results
will be correct anyways
Table 3 summarizes some results for all the lizards in the study.
[Link] Published November 2024
Page 4 of 7
Lesson
Decoding Variable Types in Data Student Handout
Table 3. Summary data for lizards at the high-elevation lab (40 lizards) and the low-elevation lab (37 lizards) over
six weeks. The change in body mass (g) and percent change in body mass were averaged for the lizards at each
lab.
Column A Column B Column C
1 Lab elevation Average change in body mass (g) Average percent change in body mass (%)
2 High −0.05 −0.56
3 Low 0.34 5.72
7. How do the results in Table 3 compare with your prediction in Question 2b?
they match because the high elevation lizards lost weight and the low elevation lizards
gained weight
PART 2: Lizard Endurance
The data in Table 3 show how high-elevation conditions, such as lower oxygen, affected the lizards’ body mass.
You’ll now explore how high elevation affected the lizards’ endurance, which was based on running
performance. Increased endurance can help lizards capture insect prey, escape predators, and pursue mates.
Moving to a higher elevation could affect a lizard’s endurance, like how some humans experience shortness of
breath at high elevations.
To measure endurance, the researchers used a brush to chase the lizards back and forth on a track, as shown in
this video. They recorded endurance as the number of seconds it took for each lizard to stop running, even
when nudged by the brush. Like with body mass, they measured the lizards at three timepoints (T1, T2, and T3).
8. Consider this research question: “Does lizard endurance differ over time between the low- and high-
elevation labs?” Generate a prediction for the result you expect.
lizards at higher elevation will have less endurance because they have less oxygen
We also might want to know if larger lizards have greater endurance regardless of elevation. One way to
perform this analysis is to first categorize the lizards as “small” or “large” based on their body mass.
9. Consider this additional research question: “Regardless of elevation, does endurance differ over time
between small and large lizards?” Generate a prediction for the result you expect.
larger lizards have larger lungs and stronger muscles so they will have greater endurance
There are multiple ways to categorize lizards based on size. For example, you could pick a specific body mass as
the cutoff between large and small lizards. Lizards greater than or equal to this cutoff could be categorized as
“large,” and lizards smaller than the cutoff could be categorized as “small.”
[Link] Published November 2024
Page 5 of 7
Lesson
Decoding Variable Types in Data Student Handout
10. Categorize each of the lizards in Table 4 as “small” or “large” based on their body mass at T1. For this
example, use 6.0 g as the cutoff between large and small lizards. Record your answers in Column D of the
table.
Table 4. Data for four lizards. Columns A–C are from the data in Table 1. As in Table 1, T1 = the first day of
the experiment.
Column A Column B Column C Column D
1 Lizard ID Lab elevation Body mass at T1 (g) Size category
2 1123 High 7.18 large
3 232 High 5.01 small
4 246 Low 7.79 large
5 245 Low 5.52 small
Table 5 shows endurance results for the lizards based on the new “Size category” variable.
Table 5. Summary data for lizards at the high-elevation lab (11 lizards) and the low-elevation lab (13 lizards) over
six weeks. Lizards were divided into size categories using 6.0 g as the cutoff between “large” and “small.” The
percent change in endurance between T1 and T3 was averaged for the lizards in each group. A percentage
change over 100% indicates that endurance more than doubled.
Column A Column B Column C
Lab Size
1 elevation category
2 High Small 136.4%
3 High Large 72.0%
4 Low Small 174.4%
5 Low Large 140.8%
11. How do the results in Table 5 compare with your prediction in Question 9?
partly the same because all of the lizards improved endurance, but the small ones showed a
higher percent increase
12. Was there a difference in endurance between lizards at the low- and high-elevation labs, regardless of size
category?
yes there was. endurance increased more at low-elevation lab than high-elevation, proving
that the availability of oxygen affects the lizards performance
13. So far, you’ve explored how lizards respond to high elevations in terms of two physiological variables: body
mass (Part 1) and endurance (Part 2).
a. Suggest another variable that could be measured to learn more about how lizards respond to high
elevations.
breathing rate
b. Describe how your variable could be measured.
[Link] Published November 2024
Page 6 of 7
Lesson
Decoding Variable Types in Data Student Handout
c. Based on your description, is your variable categorical or numerical?
d. List the specific type for your variable (nominal, ordinal, binary, discrete, or continuous).
PART 3: Design a Future Study
Now it’s your turn to come up with a study idea. Design a hypothetical study that investigates how different
elevations may affect the physiology of organisms, following these criteria:
● You will investigate the physiological effects of low, medium, and high elevations. (The “medium” elevation
will allow you to test an intermediate elevation, in addition to the extremes.)
● You can use any organism except lizards.
● In addition to elevation, you should measure two variables that could provide useful information about the
organisms in your study.
o These variables must have different types (for example, nominal and continuous).
o At least one variable should be a physiological variable (related to health or physical performance).
o Don’t use the same variables as in the lizard study (body mass or endurance).
14. Write the research question for your study.
how does elevation affect the heart rate and skin coloration of frogs living at different heights?
15. Briefly describe the design of your study, including the organisms you will use and the location of the study.
take as samples 3 types of frogs from different altitudes and
measure their heart rate and skin color
16. Fill out the following table for the variables in your study. The first variable, elevation, is included as an
example.
Variable How will the variable be Variable type and justification
measured?
Elevation Categorized as low, medium, or Ordinal categorical, because
high there are three categories that
go in a specific order.
Write your first variable beats per minute using a numerical because the
(must be a physiological sensor number of beats is a
variable): number
Write your second variable skin color, on a scale 1 to categorical because the
(must have a different type 10 where 1 is light and 10 color is a category
from your first variable): is dark
17. Generate a prediction about the difference you would expect in your two variables based on elevation.
at higher elevation since temperature is lower they will have a darker color to absorb heat and
fast heart rate because of the low oxygen
[Link] Published November 2024
Page 7 of 7