Skip to main content
Log in

Effects of emotion-induced self-focused attention on item and source memory

  • Original Paper
  • Published:
Motivation and Emotion Aims and scope Submit manuscript

Abstract

Affective states are closely linked to attention to internal aspects of the self (i.e., self-focused attention). We investigated how self-focused attention induced by emotional experiences affects memory for subsequently presented information. Prior to incidental encoding of affectively neutral target words, participants were induced to feel shame or anger through autobiographical recall (vs. no emotion-induction control condition). Memory for words (item memory) and their associated contextual features (source memory) were subsequently assessed. Self-focused attention, measured by the private self-consciousness scale, was highest in the shame condition, followed by the anger and then control conditions. Item memory was significantly impaired in the shame condition compared to both the anger and control conditions, and self-focused attention negatively mediated the effect of emotion condition on memory performance. Source memory did not significantly differ across the emotion conditions, and we discuss possible factors contributing to this null finding. Our findings suggest that emotion-induced self-focused attention may reduce attentional resources available for encoding task-relevant external information.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2

Similar content being viewed by others

Notes

  1. Source memory can be assessed in conjunction with item memory whereby participants are first asked to determine whether or not a given item had been previously presented in the encoding phase, and then, only for the items that were determined as having been presented, to further indicate their associated source feature (e.g., Doerksen and Shimamura 2001). Alternatively, source memory can be assessed independently of item memory by having participants indicate the source feature of all studied items, for example using a forced-choice test (e.g., Davidson et al. 2006). In the present study, we opted to use an independent test of source memory that is not contingent on correct item recognition.

  2. We asked participants to complete these questionnaires at the end of the experiment in an attempt to reduce the likelihood that they would be aware of the fact that their mood and the resulting self-focused attention were being manipulated or the purpose/hypothesis of the study.

  3. A parallel set of analyses using d-prime (d′) as the dependent measure produced exactly the same pattern of results. Complete statistical analyses and results are presented in Appendix 2.

  4. The same null results were obtained when source memory accuracy was conditionalised on correct item recognition (i.e., the proportion of correctly recognised old items that were attributed to their correct source, P(source correct | hit)). Complete statistical analyses and results are presented in Appendix 3.

References

Download references

Acknowledgements

This work was supported by a Grant in Support of Scholarship (GISOS) from Wesleyan University.

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Kyungmi Kim.

Ethics declarations

Conflict of interest

The authors declare that they have no conflict of interest.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Appendices

Appendix 1

The structure of the MCQ composite variables, Cronbach’s Alphas, and corresponding items.

Composite variable

Cronbach’s α

Items

Clarity

.835

My memory for the event was dim vs. sharp/clear.

My memory for the event involved visual details.

My memory for the event was sketchy vs. very detailed.

The order of events in my memory was confusing vs. comprehensible.

Overall, my memory for the event was vague vs. very vivid.

Overall, I remembered the event hardly vs. very well.

Sensory

.605

My memory for the event involved sound.

My memory for the event involved smell.

My memory for the event involved taste.

My memory for the event involved touch.

Contextual

.630

My memory for the location where the event took place was vague vs. clear/distinct.

In my memory, relative spatial arrangement of objects was vague vs. clear/distinct.

In my memory, relative spatial arrangement of people was vague vs. clear/distinct.

Time

.729

My memory for the time when the event took place was vague vs. clear/distinct.

My memory for the year when the event took place was vague vs. clear/distinct.

My memory for the season when the event took place was vague vs. clear/distinct.

My memory for the day when the event took place was vague vs. clear/distinct.

My memory for the hour when the event took place was vague vs. clear/distinct.

Thoughts and feelings

.645

I remembered what I thought at the time when the event took place: Not at all vs. clearly.

I remembered how I felt at the time when the event took place: not at all vs. clearly.

Valence of feelingsa

 

My feelings at the time when the event took place were negative vs. positive.

Intensity of feelingsa

 

My feelings at the time when the event took place were not intense vs. very intense.

  1. aThe valence of feelings and the intensity of feelings were assessed using single items

Appendix 2

Statistical results of item memory using d-prime as a measure of performance

The proportion of missing responses (Shame: M = .004, SD = .007; Anger: M = .004, SD = .009; Control: M = .004, SD = .007) did not significantly differ across the emotion conditions, F(2, 237) = 0.122, p = .885. Missing responses were counted as incorrect responses. For each participant, d-prime score was calculated by subtracting z-score-transformed false-alarm rates (the proportion of “new” words incorrectly identified as old) from z-score-transformed hit rates (i.e., the proportion of “old” words correctly recognised as old). One-sample t-tests showed that d-prime scores were significantly above chance performance level of zero across all Emotion conditions, all t(79)s > 15.472, all ps < .05. A one-way ANOVA conducted on d-prime scores with Emotion (shame, anger, control) as the between-subjects factor revealed a significant effect of Emotion, F(2, 237) = 5.317, p = .006, ηp2 = .043. Bonferroni-corrected post-hoc tests revealed that item memory was significantly impaired in the Shame condition (M = 0.770, SD = 0.445) compared to both the Anger (M = 0.973, SD = 0.372), p = .010, and Control conditions (M = 0.950, SD = 0.472), p = .026. Item memory did not significantly differ between the Anger and Control conditions, p = .999.

To examine whether self-focused attention mediated the effect of Emotion condition on item memory performance, we ran two mediation analyses with PROCESS macro for SPSS (model 4; Hayes 2018) using bootstrapping procedures with 10,000 samples. In the first analysis, we dummy-coded the Emotion condition to examine the relative direct effect, indirect effect, and total effect (i.e., the sum of the direct and indirect effects) of the Shame condition and Anger condition, respectively, relative to the Control condition (i.e., reference group). In the second analysis, we used two orthogonal contrasts to examine relative direct, indirect, and total effects of (a) Shame and Anger conditions collectively (i.e., Emotions) relative to the Control condition (contrast coded as Shame = 1/3, Anger = 1/3, Control = – 2/3) and (b) the Shame condition relative to the Anger condition (contrast coded as Shame = 1/2, Anger = – 1/2, Control = 0), respectively.

As shown in Table 5, relative to the Control condition, the Shame condition had a significant negative indirect effect on item memory accuracy via self-focused attention as well as a significant negative total effect, but a nonsignificant direct effect. Relative to the Control condition, the Anger condition had a significant negative indirect effect on item memory via self-focused attention but a nonsignificant total effect, suggesting that the nonsignificant yet positive direct effect counteracted the negative indirect effect. Relative to the Control condition, the two emotion conditions (Shame and Anger) collectively had a significant negative indirect effect on item memory accuracy via self-focused attention, but nonsignificant direct and total effects. Relative to the Anger condition, the Shame condition had a significant indirect effect on item memory accuracy via self-focused attention as well as significant direct and total effects, all of which were in a negative direction.

Table 5 Regression coefficients for mediation analyses examining the effect of Emotion condition on item memory accuracy (as calculated as d-prime score) via self-focused attention

Appendix 3

Statistical results of source memory accuracy conditionalised on correct item recognition

Source memory accuracy was calculated as the proportion of correctly recognized old words that were attributed to the correct source, P(source correct | hit). The overall proportion of missing responses (Shame: M = .004, SD = .009; Anger: M = .005, SD = .010; Control: M = .004, SD = .009) did not significantly differ across the emotion conditions, F(2, 237) = 0.246, p = .782. Missing responses were counted as incorrect responses, conditionalised on correct item recognition. One-sample t-tests showed that source memory accuracy was significantly above chance performance level of .50 across all Emotion conditions, all t(79)s > 2.884, all ps < .05. A one-way ANOVA with Emotion (shame, anger, control) as the between-subjects factor revealed no significant effect of Emotion, F(2, 237) = 0.132, p = .877 (Shame: M = .533, SD = .093; Anger: M = .531, SD = .097; Control: M = .539, SD = .104)

To examine whether self-focused attention mediated the effect of Emotion condition on source memory performance, we ran two mediation analyses using PROCESS macro for SPSS (model 4; Hayes 2018) using bootstrapping procedures with 10,000 samples. In the first analysis, we dummy-coded the Emotion condition to examine the relative direct effect, indirect effect, and total effect (i.e., the sum of the direct and indirect effects) of the Shame condition and Anger condition, respectively, relative to the Control condition (i.e., reference group). In the second analysis, we used orthogonal contrasts to examine relative direct, indirect, and total effects of (a) Shame and Anger conditions collectively (i.e., Emotions) relative to the Control condition (contrast coded as Shame = 1/3, Anger = 1/3, Control = – 2/3) and (b) the Shame condition relative to the Anger condition (contrast coded as Shame = 1/2, Anger = – 1/2, Control = 0), respectively. As shown in Table 6, the results of these analyses revealed that none of the direct, indirect, or total effects was statistically significant.

Table 6 Regression coefficients for mediation analyses examining the effect of Emotion condition on source memory accuracy (conditionalised on correct item recognition) via self-focused attention

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Jeon, Y.A., Resnik, S.N., Feder, G.I. et al. Effects of emotion-induced self-focused attention on item and source memory. Motiv Emot 44, 719–737 (2020). https://doi.org/10.1007/s11031-020-09830-w

Download citation

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s11031-020-09830-w

Keywords

Navigation