Using graphs to present data

1
Using graphs to present data

• Constructing graphs and interpreting data

2
Types of graphs commonly used in social work

• Bar charts
• Pie charts
• Histograms
• Line charts
• Charts for single system designs

3
You must use graphs that are appropriate to the
level of measurement associated with the variable
you are measuring
Type of Chart Level of Measurement
Bar Chart Nominal must be organized into categories
Pie Chart Nominal, ordinal, interval, or ratio. However, it is not practical to use a pie chart when there are more than five or six possible values for a variable.
Histogram Ordinal, interval, or ratio level data. Most often used with ratio or interval level data
Line Chart/Frequency Polygram Interval and ratio data
Single System Design Interval and ratio data
4
Bar Chart Number of children with income under
the poverty line
5
Pie Chart
6
Percents for Pie Chart
African American American Indian Asian Caucasian Latino Other Total
5 2 10 20 25 8 70
7.14 2.86 14.29 28.57 35.71 11.43 100.00
7
Histogram Ages of MSW students responding to
exit survey
Graph
8
Line Chart
9
Techniques for Making Comparisons Histogram
10
Techniques for Making Comparisons Line Chart
11
Single System Designs

• Used by practitioners to measure whether
  intervention is effective.
• Not a formal evaluation of worker performance.
• Comparisons are made by comparing baseline
  measures to intervention phase.
• Measures used are usually ratio standardized
  scores or counting behaviors.
• Behavioral counts are usually self-reports
• Sometimes comparisons are made across clients,
  types of interventions, or types of behaviors

12
Single system Design Graph
13
Single System Designs include

• At least two phases, baseline and intervention.
  (AB design)
• Baseline is a period when clients do not receive
  an intervention.
• Intervention is when treatment is offered.
• The reason this is done is because things may be
  happening in the clients life that affects the
  target behavior you want to control for these
  effects by alternating no treatment with
  treatment you can see if patterns are similar
  during both phases.
• Intervention is successful if there are changes
  in the trend line (up or down) after the
  intervention starts

14
Other types of designs

• ABA (baseline, intervention, baseline.
• ABC (baseline, Intervention 1, Intervention 2,
  etc.)
• Multisystems designs targets two or more
  behaviors in the same client or two or more
  clients with the same problems and target
  behavior or attribute.
• Baseline for client number two is extended until
  after the intervention has started for client 1
  in order to further control for the effects of
  the surrounding environment.

15
Issues with Single System Designs

• A variety of factors can influence the success or
  failure of the intervention the agency, the
  worker, client-worker interaction, the clients
  home or work environment what happens in the
  clients community or the country.
• You cant generalize findings from one single
  system design to other clients, workers or
  settings.