To Count or Not to Count: That is the Instruction Danerys Encarnacion, Suzanne StahlBell and Mark S.

1
To Count or Not to Count That is the
InstructionDanerys Encarnacion,Suzanne
Stahl-Bell andMark S. Schmidt

• Encarnacion, D., Stahl_Bell, S., Schmidt, M.S.
  (2002, April). To count or not to count. Paper
  presented at the 1st annual Columbus State
  University Student Colloquium, Columbus, GA.

2
Program Overview

• Early Research
• Hypothesis
• Method
• Results
• Conclusion

3

• Enumeration tasks measure the ability of people
  to perceive the number of targets presented in a
  visual display.

4
Early Research

• Kaufman et al.
• - 1-210 targets presented for 200 ms
• - Found that response time leveled off and
  accuracy greatly dropped when presented with
  more than 6 targets.
• - Conclusion 2 different process
• 1) subitizing (1-6)
• 2) estimation (gt than 6)

5
Estimation
subitizing
6
Early Research (cont.)

• Dehaene and Cohen
• - Found that people with simultanagnosia could
  not count 4-6 dots, but they could subitize 1-3
  in a task.
• - Conclusion 2 separate processes occur in
  what Kaufman just called subitizing.
• 1) subitizing (1-3)
• 2) counting (4-6)

7
Current Theory

• There are 3 different process occurring during an
  enumeration task (Mandler and Shebo)
• 1) Subitizing 1-3 targets
• -fast and accurate
• 2) Counting 4-6 targets
• -slower and less accurate
• 3) Estimating 7-9 targets
• -slower and inaccurate

8
Counting from WM
Estimation
Subitizing
9
In Contrast to Previous Results

• Thomas et al.
• -Two choice task was used (4 vs. 5, 6 vs. 7, 8
  vs. 9, etc) for 3-11 dots
• -Participants were told not to count
• Results Response times staying in the same
  range without reduced accuracy (90 correct)
• Conclusion Conceptual Prototype Matching
• Replaces subitizing, counting and estimation.

10
Prototype Matching

• The ability to rapidly recognize an example of a
  concept
• For example, being able to recognize a bird, a
  tree, a house, etc.
• Maybe people can learn to recognize the number of
  things without counting in the same way
• Ex) 8-ness, 7-ness, 9-ness, etc.

11
(No Transcript)
12
(No Transcript)
13
Our Hypothesis

• 1) Response times for participants instructed
  Not to Count would be quicker than response
  times for participants instructed to count.
• 2) The control groups responses would be in
  between.

14
Method

• Participants 65 Undergraduate Students
• Apparatus
• - IBM compatible 486 computer
• - Standard computer monitor
• - Computer program written by Dr. Schmidt
• - Cassette tape recorder
• - Standard computer microphone

15
Procedure

• 3 Instruction conditions
• 1) Try to count
• 2) Dont try to count
• 3) Control (not instructed either way)

16
Procedure (cont.)

• Order of events on each trial
• Fixation point for 1 sec
• Display of dots for 200 ms
• Masking stimulus until response was made

17
(No Transcript)
18
(No Transcript)
19
(No Transcript)
20

• An additional 5 participants were tested with
  Arabic numerals to make sure the voice key was
  responding equally to all verbal pronunciations
  of the numbers.

21
(No Transcript)
22
Results

• Arabic numbers no significant differences
  between the groups on response time

23
(No Transcript)
24
Results

• Response times Dont count condition was faster
  than count condition when presented with 6 or
  more dots.

25
(No Transcript)
26
Results

• Accuracy no significant differences between the
  groups.

27
(No Transcript)
28
Conclusion

• Response times for the Dont Count condition
  were faster but just as accurate as the response
  times for the Count condition.
• No speed-accuracy tradeoff.
• Participants might have been using the fast
  prototype matching which Thomas et al. (1999)
  suggested was used in their two-choice tasks.

29
Thank You
30
(No Transcript)
31
(No Transcript)
32
(No Transcript)
33
(No Transcript)
34
(No Transcript)
35
(No Transcript)
36
(No Transcript)
37
(No Transcript)
38
(No Transcript)