Data Collection Methods

(Adapted from: Olswang & Bain, 1994)

Quantitative Data / Qualitative Data
Description / Quantitative data are numerical. The numbers can represent actual values or arbitrary categories (e.g., rating scale). / Qualitative data are verbal. They are often written notes that are
descriptive and/or reflective.
Types /
  • Event/Frequency Recording
  • “Did it happen?”
  • (+/-; % occurrence)
  • Duration Recording
  • “For how long did it happen?”
  • (sec., min.)
  • Interval Recording
  • “Did it happen during this time?”
  • (+/-; % occurrence per time)
/
  • Observation notes (“Field notes”)
  • Interview notes
  • Document review
  • Photograph review

Examples / 70% (14/20) accurate production
Average of 3 min. on-topic exchange
Appropriate social interaction during 9/10 5-minute intervals on playground / “Student responses were tangential. He did not respond to my non-verbal cues.”
“Student had a meltdown when the teacher asked her to clean. My impression was that the transition was not planned in advance and was unclear to the student.”

Types of Data to Support Practice-Based Evidence

(Adapted from: Olswang & Bain, 1994)

Session Data / Generalization Data / Control Data
Type of Question / Is the student responding to treatment? / Is there evidence of significant and important improvement? / Is there evidence that my treatment is responsible for this improvement?
Description / Data that reflect the student’s performance while completing targeted activities during treatment sessions. / Data that reflect how the student performs targeted skills outside of treatment. Can be measured in several ways:
  • untrained stimuli
  • novel environments
  • novel partners
/ Data that reflect that a student is NOT changing on untreated behaviors without intervention.
Frequency / Ongoing
(Record data during each session) / Probe
(Occasional sampling of performance outside of treatment) / Probe
(Occasional sampling of performance on untreated behaviors)
Hypothesized Changes
Post-Treatment / Consistent, clear improvements over baseline / Depends on intervention theory:
  • If expect generalization, want to see improvements.
  • If not, don’t need to see improvements.
/ Should show no change since baseline. *If you do show improvements on control data, you must be wary of interpreting the value of your treatment. Something else may have caused the student to improve (e.g., age, awareness, classroom).
Conclusion / The student improved in therapy. / The student improved in therapy AND this made an impact on functional, academic, real-life, everyday behavior. / The student has demonstrated significant and important improvements on everyday, functional tasks as a result of this intervention.