STUDIES - Tversky and Kahneman (1974)

Question 1

Aim

Answer 1

• To demonstrate how the anchoring bias affects students when solving a maths problem

Question 2

Procedure

Answer 2

SAMPLE

• High school students

STEPS
Participants were allocated to 2 groups:

• Group 1 (ascending condition): Participants were asked to estimate the value of 1x2x3x4x5x6x7x8 in 5 seconds
• Group 2 (descending condition): Participants were asked to estimate the value of 8x7x6x5x4x3x2x1 in 5 seconds

Question 3

Results

Answer 3

• Group 1 (ascending condition): Median was 512
• Group 2 (descending condition): Median was 2250
• Actual value: 40320
• Group 2 had a median 4 times higher than group 1

Question 4

Findings

Answer 4

• Since we read left to right, group 1 used “1” as an anchor and estimated a lower value than the group that started with “8” as the anchor
• First number seen biases the estimate of the value by participant

Question 5

Strengths

Answer 5

• High replicability: Very simple experiment that allows us to establish the reliability of the results
• High internal validity: The study was highly controlled as results were quantitative and easy to calculate, and it can be inferred that anchor was cause of estimates by students (cause-and-effect relationship)
• Use of median: Researchers used median to report the data, allowing them to diminish the influence of outliers

Question 6

Limitations

Answer 6

• Low ecological validity: The situation is very artificial, as it is unrealistic to be given only 5 seconds to estimate the value of something
• Application of findings: Questionable to what extent they can be applied
• Low internal validity: Participant variability may have influenced the results due to the design used (independent samples), it would have been better to use matched pairs design to have 2 groups with an equal level of maths