Printer Friendly

Activity: Data Representation

 

Materials

Data
Computer with graphing program or graph paper, ruler, and pencil

Procedure

  1. Write down the initial research question or hypothesis that the data set was meant to answer or test.
  2. Write down who your audience is, specifically, to whom are you presenting your data and why?
  3. Use your knowledge of your audience and your research goal or hypothesis to select data representation method(s) from Table 1. Choose carefully because you will need to be able to defend your method(s) of displaying your data. If you choose a graph, look at Table 2 to find more detailed information about the type of graph to use.
 
Table 1: Data representation methods

Type of Data Representation

Advantages

Disadvantages

Raw Data

(Tables)

· Shows full record

 

· Allows comparisons (via sorting by variable)

 

· Can be unclear

 

· Hard to make inferences

Narrative

(Oral or Written)

· Descriptive

 

· Conveys meaning

· Presentation can take a long time (orally) or be bulky (written)

Descriptive Statistics

(mean, median, mode,

range, standard deviation)

· Summarizes

 

· Reveals correlations

· Outliers can influence statistics

 

· Can be misleading or obscure information about the data set

Graphs

(see Table 2)

· Visual

 

· Summarizes

· Can be misleading

 

· Can be confusing

 

 

Table 2: Types of data appropriate for different types of graphs and cautionary notes

Graphs*

Type(s) of Data

Notes

 

Pie

· Percents of population

 

· Percent abundance

· Must add to 100%

 

· Use only for data that add up to meaningful total

 

· Avoid forcing comparisons across more than two pie charts

 

Bar

· Abundance/quantity comparison data

 

· Categorical data (usually on x-axis)

· Area of bars can imply volume

 

· Categorical names with long x-axis labels can be moved to the y-axis

 

Scatter

· Numerical data (both axes)

 

· Two or more variables

 

· Look for patterns

· Best fit line or regression line can be used to show trends

· Only extend line to axis if meaningful

 

Line

 

· All data points are connected in a logical sequence (e.g. time series or distance)

 

· The line connecting points is not based on statistics or patterns. All points are connected.

· Connect points with a line only when appropriate

 

· Can be used for categorical data

 

Table of Contents:

Exploring Our Fluid Earth, a product of the Curriculum Research & Development Group (CRDG), College of Education. University of Hawaii, 2011. This document may be freely reproduced and distributed for non-profit educational purposes.