go to the School of Psychology home page Go to the UNE home page
Chapter 7 - Analysing the Data Part IV - Analysis of Variance Chapter 1 - Behavioural Science and research Chapter 2 - Research Design Chapter 3 - Collecting the Data Chapter 4 - Analysing the Data Part I - Descriptive Statistics Chapter 5 - Analysing the Data Part II - Inferential Statistics Chapter 6 - Analysing the Data Part III - Common Statistical Tests

 

Chapter 7: Analysing the Data
Part IV : Analysis of Variance

 

Scenario and Data Set #2
Report

Task 1.

Table 1. Means, standard deviations, and sample sizes for each experimental group.

Experimental groups

Counting

Rhyming

Adjective

Imagery

Intentional

Mean words recalled

7.0

6.9

11.0

13.4

12.0

sd

1.83

2.13

2.49

4.50

3.74

N

10

10

10

10

10

Task 2.

The complete summary table. In the present case, the Summary table displayed by SPSS is exactly the one we want. In higher order ANOVAs this is not always so.

Summary Table

Source of variation

Sum of Squares

df

Mean Square

F

Sig.

Between Groups

351.520

4

87.880

9.085

.000

Within Groups

435.300

45

9.673

Total

786.820

49

Task 3.

Levene's test for homogeneity showed that the assumption of equality of variance among the five groups was not violated. The analysis results can therefore be considered valid.

Task 4.

There was a significant difference amongst the means of the five experimental groups (F(4,45) = 9.09, p < .001). Follow up Tukey's HSD tests indicated that the two groups with the lowest level of processing (Counting and Rhyming) showed the lowest level of recall. The Adjective, Imagery, and Intentional groups all showed significantly greater recall than the Counting and Rhyming groups. The Intentional group did not show any more recall than the Adjective or Imagery groups.

Task 5.

Figure 7.7. SPSS graph of mean number of words recalled

This is the default graph that SPSS gives as a result of checking "Means Plot" in Figure 7.5. It can be considerably improved by editing (double clicking on the image in the Output Viewer) and changing a few things. In particular, because we dealing with discrete groups here that do not follow any obvious continuum, the graph should be changed to a bar graph.

An improved graph is shown below:

Task 6.

The data provide some support for the model of memory as proposed by Craik and Lockhart (1972). The degree to which verbal material is remembered depends to some extent on the level of processing required when the material was originally presented. Here the two groups exposed to the lowest levels of processing recalled fewer words than the three groups exposed to higher levels of processing. Learning therefore depends on more than mere exposure to the material.

The results suggest a threshold in the optimum level of processing such that below this threshold word recall is lower than if processing is above the threshold. This threshold occurs at the point when a person has to pay attention to the meaning of the whole word. If a person treats each word as a collection of letters then recall is poor. If, however, the word has to be processed for meaning (regardless of whether the meaning is for imagery or an adjective), then recall is higher. Instructions that indicate that the person would be later tested for recall had no effect over and above that required to process the whole meaning of the word.

The reading of newspaper articles would involve whole word processing. Processing of incidental advertisements would be more like the lower levels. Recall should therefore be better for the articles than the advertisements!

 

 

 

© Copyright 2000 University of New England, Armidale, NSW, 2351. All rights reserved

UNE homepage Maintained by Dr Ian Price
Email: iprice@turing.une.edu.au