Hypothesis

It was hypothesized that words processed via self-referent encoding will be remembered better than words that are processed via structural encoding.

Method

Participants

            The experiment involved participants from the University of Memphis (n = 36). The participants completed experiment on the APA Online Psychology Laboratory. The age range of the participants was 19 to 43 years with the average being 23.00 years with a standard deviation of 6.13. Males were 5 while females were 30 and 1 of the participants was gender nonconforming. There was no compensation for participation in the experiment.

Materials

The experiment consisted of two phases; the encoding phase and the retrieval phase. For the encoding phase, 20 common target words were used. Half of them had an ‘e’ in them and half of them were less socially desirable. The encoding questions asked whether the target word had an ‘e’ or whether the target word was self-descriptive. The same list of 20 target words was used in the retrieval phase with another 20 common distractor words. All this was done on a computer.

Procedure

            Before the commencement of the experiment, the participants were given a brief description of the study and then asked to read and sign an informed consent form. The participants then authenticated to press continue or quit. For those who decided to continue with the experiment, they were taken to the first phase of the experiment, the encoding phase. In this phase, the participants were presented with 20 common words one after another. The order of presentation was random for each participant. Following the presentation of each word, participants were asked one of the encoding questions, i.e., does word contain a “e” –OR– does this word describe you? Half of the words were randomly assigned to each question. To ensure an equal number of yes and no responses, half of the target words had an ‘e’ in them and half the target words were less socially desirable. Each word was displayed for 1 second and the participant had only 1 second to choose between a ‘yes’ and ‘no’.

After a short delay, participants were given a recognition test, which was the retrieval phase. The participants were informed that they would be shown some new words and some words that they had seen earlier when answering the yes or no questions. Their task was to indicate which of the words were the ones they saw earlier and which ones were new. For instance, they could be presented with the word ‘clumsy’ and asked: ‘was this one of the words you saw earlier?’ The answer was yes or no. The data was then recorded for analysis.

Design

The independent variable (IV) was the level of encoding, which had two levels including self-referent and structural. The dependent variable (DV) was the accuracy of identifying original target words. Accuracy was a composite measure called A-prime (A’) that included both hits and false alarms.

A’ was calculated for both self-referent and structural encoded words using the following formula: A’ =  + . A’ for each level of the IV was analyzed via a paired-samples t-test.

Results

The hypothesis for experiment was that when self-reference is used to encode target words they will be remembered more accurately than when word structure is used to encode target words. This hypothesis was supported indicating that accuracy for the self-referent condition was significantly higher than the accuracy for the structural condition, t (35) = 6.41, p < .001. As shown in figure 1, the mean accuracy for self-referent condition (0.96), which was group 1, was higher than the mean accuracy for the structural condition (0.72), which was group 2.

Figure 1: Descriptive and Inferential Statistics from the paired-samples t-test

Discussion

The experiment was done to determine if self-referent information is remembered better than structural information. The results of the experiment revealed that words processed via self-referent encoding will be remembered better than words that are processed via structural encoding. The p value from the t-test was less than .001 thereby indicating a significant difference in the means of the two encoding conditions. This was visually supported in figure 2, which shows that the mean discrimination index for the self-referent condition was significantly higher than the mean discrimination index for the structural condition. Also, the figure shows aa lesser standard error for the self-referent condition than the structural condition. The results of this experiment support the claims made by Rogers, Kuiper, and Kirker (1977) on the tendency to remember words that have personal relevance more than those requiring semantic processing. This could serve the greater purpose of manipulating the psychology of students during teaching by using words that are self-referent to the students to harness their memory capability.

Figure 2: Comparison of the mean discrimination index for the two encoding conditions.