Experiment 7: Immediate Serial Recall Task

The outcomes of Experiment 3 indicated that there were no correlations between let-ter-crossing measures and ISR task performance save for 4-syllable items and letlet-ter-crossing accuracy. As such, there was no strong expectation of group differences on any of the

measures in the current experiment. Experiment 3 showed a significant word length effect in that 2-syllable words were better recalled than 3-syllable words, which were better recalled than 4-syllable words. Repetition effects are expected to be observed, with an increase in per-formance expected between pre- and post-test. If the results of the previous experiments are to be replicated, there should be no group by session interactions with any of the measures, save for the distractor measure. Experiment 7 reported strong power, N = 109, α = .05, β = .01, Power = .99, with a critical F value of 3.08.

6.6.1. Methods.

The methods are the same as those employed in Experiment 3, Study 2. The partici-pants, materials, and procedure were fully described in Experiment 3.

6.6.2. Results.

The analyses conducted were 3 (Group: Higher, Moderate, and Lower performance) x 3 (Syllables: 2-Syllable, 3-Syllable, and 4-Syllable items) x 2 (Session: Pre-test, Post-test) mixed design ANOVAs for letter-crossing accuracy, speed, and slope group splits. A 2 (Group: Distractors, No-Distractors) x 3 (Syllables: 2-Syllable, 3-Syllable, and 4-Syllable items) x 2 (Session: Pre-test, Post-test) mixed design ANOVA was conducted for the distrac-tor group split. Word length effects were found with 2-syllable words recalled more often than 3-syllable words, which were better recalled than 4-syllable words. Post-test recall was also superior to pre-test performance. Group differences appeared to be weaker in this experi-ment. Where a variable violated Mauchley’s test of sphericity, the Huynh-Feldt correction was used, as ɛ > 0.75. The outcomes of the study are summarised in Table 6.3 and Figure 6.3.

Table 6.3

Mean and Standard Errors For Letter-Crossing Accuracy, Speed of Completion in Minutes, Slope of Accuracy Over Time, and Distractor Recalls Split By Letter-Crossing Groups (Higher, Moderate, and Lower) Across Pre- and Post-Test Sessions in 2-, 3-, and 4-Syllable Items Under the ISR Task

ISR Task

Letter-Crossing Task

2-Syllables 3-Syllables 4-Syllables

Pre Post Pre Post Pre Post Distractors No-Distractors (N =

74)

2-Syllables 3-Syllables 4-Syllables

Figure 6.3. Mean accuracy (y axis) scores across pre- and post-test sessions (x axis) for ISR syllable lengths: 2-syllables, 3-syllables, and 4-syllables. Error bars were created from standard error values. Whereby Higher performance refers to Higher Accuracy, Faster Speed, Minor

2-Syllables 3-Syllables 4-Syllables

Slope, and No-Distractors; Moderate performance refers to Moderate Accuracy, Moderate Speed, Moderate Slope; and Lower performance refers to Lower Accuracy, Slower Speed, Major Slope, and Distractors.

6.6.2.1. Accuracy split. There was no significant difference between groups, F (2, 106) = 2.44, p = .092, MSE = .13, partial η² = .04. A main effect was found for session, F (1, 106) = 55.32, p <.001, MSE = .02, partial η² = .34. Significant improvements were seen from pre- to post-test. There was no interaction between session and accuracy group, F (2, 106) = 55.32, p = .55, MSE = .02, partial η² = .01. A main effect for syllable-length was identified, F (2, 212) = 171.38, p <.001, MSE = .01, partial η² = .62. Where higher averages were reported under 2-syllable items, followed by 3-syllable and 4-syllable items. The ses-sion and syllable interaction was significant, F (1.92, 203.76) = 3.29, p = .041, MSE = .01, partial η² = .03. Planned contrasts revealed a significant interaction, F (1, 106) = 5.54, p = .020, where a significant increase was observed for all syllable lengths from pre-test to post-test. No other interactions were found.

6.6.2.2. Speed split. There was a significant difference found between groups,

F (2, 106) = 3.49, p = .034, MSE = .12, partial η² = .06, where the faster speed group consist-ently reported greater accuracy compared to the moderate and slower speed groups respec-tively. A main effect for session was found, F (1, 106) = 55.78, p <.001, MSE = .02, partial η² = .35. Significant improvements in target recalls were seen from pre- to post-test. There was no interaction between session and speed groups, F (2, 106) = 1.55, p = .218, MSE = .02, partial η² = .03. A main effect for syllable was found, F (2, 212) = 180.63, p <.001, MSE = .01, partial η² = .63. Once again, the highest averages were reported under the 2-syllable items. This was followed by 3-syllable and 4-syllable item averages. The session and syllable interaction was significant, F (1.92, 203.77) = 3.19, p = .045, MSE = .01, partial η² = .03.

Planned contrasts revealed a significant linear interaction, F (1, 106) = 5.50, p = .021, where improvements were seen across all syllable lengths from pre- to post-test. No other interac-tions were found.

6.6.2.3. Slope split. There was no significant difference found between groups,

F (2, 106) = .42, p = .66, MSE = .13, partial η² =.01. A main effect for session was found, F (1, 106) = 59.16, p <.001, MSE = .02, partial η² = .36. Significant improvements were found from pre- to post-test. There was no interaction between session and slope groups, F (2, 106)

= .5, p = .609, MSE = .02, partial η² = .01. Significant differences between syllables lengths were found, F (2, 212) = 165.07, p <.001, MSE = .01, partial η² = .61. The highest average proportions were recalled for 2-syllable items. This was followed by 3-syllable items and

4-syllable items. The session and 4-syllable interaction was significant, F (1.91, 202.75) = 4.40, p

= .015, MSE = .01, partial η² = .04. Planned contrasts revealed a significant linear interac-tion, F (1, 106) = 7.53, p = .007, an increase in recall rates were observed across all syllable lengths from pre-test to post-test. No other interactions were observed.

6.6.2.4. Distractor split. There were no significant differences identified be-tween distractor groups, F (1, 107) = 3.13, p = .08, MSE = .13, partial η² = .03. There was a main effect for session, F (1, 107) = 46.69, p <.001, MSE = 1.03, partial η² = .3. A signifi-cant improvement in target recalls was seen from pre- to post-test. An interaction between session and distractor group occurred, F (1, 107) = 8.60, p = .004, MSE = .02, partial η² = .07. Planned contrasts revealed a significant linear interaction, F (1, 107) = 8.60, p = .004, where both groups improved from pre-test to post-test. There was a main effect for items, F (2, 214) = 171.14, p <.001, MSE = .01, partial η² = .62. The highest accuracy proportion av-erages were consistently reported under 2-syllable items, followed by 3-syllable and 4-sylla-ble items. There were no other interactions found.

6.6.3. Preliminary discussion.

The expected word length and repetition effects emerged in the data. As was the case in the previous experiments, there were no group by session interactions when groups were split on the basis of overall accuracy, speed, or slope values. With these measures, there is lit-tle evidence for depletion transfer effects. In contrast, post-test differences between distractor groups are much larger than pre-test differences. Thereby, the current experiment replicated a similar finding in Experiment 6, where individuals susceptible to distractor responses show limited repetition effects on the post-test, which suggests a functional marker of a depletion transfer effect was present.