SPSS Statistics

How do you perform a repeated measures ANOVA with a covariate?

Murre (2021) published a replication a famous study by Godden & Baddeley (1975) where they examined the effect of encoding/retrieval contexts. There were 4 groups of subjects with the following conditions:

Group 1)  Encoding (Wet) - Retrieval (Wet)

Group 2)  Encoding (Wet) - Retrieval (Dry)

Group 3)  Encoding (Dry) -  Retrieval (Wet)

Group 4)  Encoding (Dry) -  Retrieval (Dry)

For encoding, participants memorized words under 2 conditions (wet or dry), and for retrieval, half of each group tried to recall the words under wet and dry conditions.

Thus, dependent variable was "number of recalled words".

Originally the author used a two-way anova with encoding condition and retrieval condition as independent variables. However, in his "addendum" (Addendum to: Murre (2021). The Godden and Baddeley (1975) experiment on context-dependent memory on land and underwater: a replication | Royal Society Open Science (royalsocietypublishing.org) , he reported that the correct way to analyse was to use a repeated measures ANOVA with an order effect as a between-subjects covariate.

As I understand, repeated-measures ANOVAs require data where there are at least 2 data points for each line (participant), such as Trial 1, Trial 2. With just one data point (average number of words recalled) per subject, how can one perform a repeate-measure anova?

I would appreciate any insight

------------------------------
R Azuma
------------------------------

Hi @R Azuma 

The IBM SPSS Statistics Documentation contains Case Studies for just this purpose - explaining our procedures.  Here is a case study regarding the GLM Repeated Measures functionality.  This procedure is part of the IBM SPSS Statistics Standard Edition, or specifically the IBM SPSS Advanced Statistics module.

I hope this helps

How do you perform a repeated measures ANOVA with a covariate?

Murre (2021) published a replication a famous study by Godden & Baddeley (1975) where they examined the effect of encoding/retrieval contexts. There were 4 groups of subjects with the following conditions:

Group 1)  Encoding (Wet) - Retrieval (Wet)

Group 2)  Encoding (Wet) - Retrieval (Dry)

Group 3)  Encoding (Dry) -  Retrieval (Wet)

Group 4)  Encoding (Dry) -  Retrieval (Dry)

For encoding, participants memorized words under 2 conditions (wet or dry), and for retrieval, half of each group tried to recall the words under wet and dry conditions.

Thus, dependent variable was "number of recalled words".

Originally the author used a two-way anova with encoding condition and retrieval condition as independent variables. However, in his "addendum" (Addendum to: Murre (2021). The Godden and Baddeley (1975) experiment on context-dependent memory on land and underwater: a replication | Royal Society Open Science (royalsocietypublishing.org) , he reported that the correct way to analyse was to use a repeated measures ANOVA with an order effect as a between-subjects covariate.

As I understand, repeated-measures ANOVAs require data where there are at least 2 data points for each line (participant), such as Trial 1, Trial 2. With just one data point (average number of words recalled) per subject, how can one perform a repeate-measure anova?

I would appreciate any insight

------------------------------
R Azuma
------------------------------