A comparison of magnitude estimations and category judgments. Primary data.

Cognitive Psychology

Authors(s) / Creator(s)

Petzold, Peter
Haubensak, Gert

Abstract

These experiments assessed, for the magnitude estimations of squares, whether the range of the sequential dependencies of magnitude estimations and the categorical judgments vary. To this end, partial correlations between the assessment of a recently presented stimulus and the preceding stimuli and judgments were calculated. It was found that the range of magnitude estimations amounted to 1, while for categorical judgments this value was 2.
Furthermore, a relationship was found between the influence of preceding judgments in the sense of assimilation and the influence of preceding stimuli towards a contrast.
a) The difference in the range of sequential dependencies for the two types of judgments affects both the assimilation of judgments as well as the contrast with respect to the stimuli.
b) In the analysis of individual differences in the strength of sequential dependencies, a strong positive correlation was found between the extent of the influence of previous judgments and previous stimuli.
For categorical judgments, relationships between the sequence effects caused by stimulus-judgment events generated one and two trials prior to the current trial, were analyzed. Position effects resulted with respect to the relative position of the currently presented stimulus and the preceding stimuli one and two trials prior. If the squares presented one and two trials prior are both larger or both smaller than the current square, there is an interaction between them. However, if one of the previous two stimuli is smaller and the other is larger than the current stimulus, the interaction is eliminated.
The results obtained confirm a model of multiple standards, in which the stimuli are evaluated in categorical judgments with respect to a subjective area. The subjective area can be formed by the boundaries of the stimulus area (long-term standards) and/or memory representations of previous stimuli (short-term standards). In a trial, the two standards to which the stimulus to be judged is most similar are chosen from the set of available standards.

Persistent Identifier

https://doi.org/10.5160/psychdata.pdpr99ve20

Year of Publication

2004

Funding

German Research Foundation

Citation

Petzold, P. & Haubensak, G. (2004). A comparison of magnitude estimations and category judgments. Primary data. (Version 1.0.0) [Data and Documentation]. Trier: Research Data Center at ZPID. https://doi.org/10.5160/psychdata.pdpr99ve20

Study Description

Research Questions/Hypotheses:

The range of sequence effects is larger than the size estimates for categorical judgments.
For categorical judgments, there are interactions between the influence of the events 1 and 2 trials back.

Research Design:

Experimental Design, Mixed Design, Laboratory Experiment; single measurements

Measurement Instruments/Apparatus:

The amount of the stimuli consisted of 12 squares which differed in length. The lengths ranged from 50-83 mm:
Square 1: 50mm
Square 2: 53 mm
Square 3: 56 mm
Square 4: 59 mm
Square 5: 62 mm
Square 6: 65 mm
Square 7: 68 mm
Square 8: 71 mm
Square 9: 74 mm
Square 10: 77 mm
Square 11: 80 mm
Square 12: 83 mm
40 students who had no experience with this kind of experiment took part. Subjects were randomly divided between the 2 conditions. 31 participants performed categorical judgments (Experiment 1) and 9 estimated size (Experiment 2). The wording of the instructions is sound in the program. In Experiment 1, subjects were asked to judge the size of the squares on a 5-point rating scale. They had to press the corresponding number key on the keyboard.
Square 6 was shown at the beginning Experiment 2 along with the number 10 as a modulus. Subjects were asked to assess the following squares relative to this standard. If a square appeared twice as large as the standard square, subjects were to chose the number 20. In this experiment, the selected number was typed into the computer’s keyboard.
In both experiments, squares appeared 1 second after a response. No feedback was given. Sessions consisted of 3 parts, each lasting about 15 minutes with a 5-minute break. The squares were presented randomly. It was verified that the autocorrelation function of the stimuli up to 8 trials back adopted no significant value.

Data Collection Method:

Data collection in the presence of an experimenter

Individual administration
Computer-supported

Population:

Survey Time Period:

1997 to 1998

Sample:

Convenience Sample

Gender Distribution:

Age Distribution:

Spatial Coverage (Country/Region/City): Germany

Subject Recruitment:

Sample Size:

40 individuals

Return/Drop Out:

dpr99ve20_readme.txt

Text file - 3 KB