Background and aims
If you are at a loud cocktail party and someone says your name on the other side of the room, you will likely notice it, even though you cannot hear the rest of the conversation. This is known as the cocktail party effect [1] and is the most famous example of self-bias (prioritised processing of self-related information). Self-bias extends beyond attention to memory [2], perception [3] and decision-making [4]. For example, in the shape-label matching tasks [4], when participants are told to associate shapes with different people, they will be fastest and most accurate at judging whether a presented pairing matches one previously learnt when the pairing relates to themselves. The self-bias is usually considered extremely stable; however, mood disorders influence it.
Negative mood induction leads to reduced self-biases [5]. Further research has been conducted in naturally depressed individuals, but no change has been observed in self-bias [6,7]. However, extension of the shape-label matching task to include emotional faces has found emotional insensitivity in depressed populations [6,7]. This contrasts the self-negativity bias usually predicted by mood disorder theories [8]. This contradiction may stem from the differences in explicit and implicit expressions of depression. Currently, depression diagnosis and treatment efficacy are largely dependent on explicit expressions measured by self-reporting. The dependency on self-reporting is problematic as it often leads to underreporting [9] and may neglect important changes in cognitive functioning [10].
The aim of this feasibility study is to explore whether implicit cognitive biases (e.g., self emotion-biases) can track changes in mood and hence be a potential tool for depression diagnosis, treatment efficacy evaluation and provide models for novel interventions. To address this issue, simple, emotionally integrated identity shape-label matching tasks will be conducted twice weekly over a four-week period in both depressed and healthy populations over four age-group categories. Two additional follow ups at six and eight weeks will explore differences across larger time periods.
We will investigate (i) whether the implicit cognitive bias scores fluctuate with mood and (ii) whether changes in self emotion-bias at early time points can predict future moods. We predict that healthy adults will show greater self-bias with positively integrated shapes and reduced self-bias when negative emotions are integrated due to their positive self-concept. Based on negative mood induction studies, it is expected that on days when healthy individuals have higher negative affect, they will exhibit reduced self-biases. In depressed populations, evidence of emotional insensitivity (reduced positive AND negativity bias) is predicted. How cognitive biases will change dependent on mood in depressed populations is exploratory. Exploratory analysis will be conducted comparing the influence of mental health and age on self emotion-bias stability and magnitude. Underlying cognitive processes will also be explored.
Methodology
Sample size
Sample sizes between 24 and 50 have been recommended for feasibility studies [11,12,13]. Previous studies using similar experimental paradigms (in single sessions) suggest that a sample size of 23 is needed for a medium-sized mood modulation effect based on power analysis calculations [14]. 40 participants will be recruited in each age category (19-34: young, 35-49: early middle-aged, 50-65: late middle-aged and 66+: elderly) and condition (healthy and depressed) to allow for dropout due to the longitudinal nature of the study and elimination of participants based on preregistered exclusion criterion. The gender ratio will be representative of the real-world populations in each age category.
Prescreening
Prolific’s prescreening options will be used to reach target participants. A series of short questions (5 minutes) will be used to confirm participant suitability for each condition.
Procedure
Participants will complete two sessions each week for four weeks, then an additional session at week six and eight. In each session participants will complete the Positive and Negative Affect Schedule [15] to measure mood and the implicit cognitive bias task. The implicit cognitive bias task will use the shape-label matching task [4] with the addition of emotion cue and integration to assess self-emotion biases (Figure 1). The task is short and simple, taking 8-10 minutes. In this task, participants learn shape-label pairs (e.g., You are a pentagon). Shapes and labels are then presented, and participants judge whether they match a pairing previously learnt or not. Emotion cues are presented and then integrated within the shape. Participants are told the emotion is task irrelevant. Try the task here! In the first session of weeks 0, 2, 4, 6 and 8 participants will complete the Beck’s Depression Inventory [16] to track depressive symptoms and Ryff’s Scales of Psychological Well-Being [17] to assess wellbeing.
Figure 1. An overview of the implicit cognitive bias task.
Analysis plan
Correlational analysis will be conducted to compare self-reported mood and depressive symptoms and the implicit cognitive biases. Time course and regression analysis will be used to analyse the changes in mood, bias scores and the use of these scores for predicting future scores. Generalized linear mixed models will be used to analyse the reaction time, accuracy and relative bias data [18,19]. Models with fixed effects to test the self-prioritisation effect, emotion biases, interactions of person with emotion cues, positive and negative affect and depressive symptoms will be conducted. Additionally, hierarchical drift diffusion models will be used to explore underlying cognitive processes.
Study costs
*Up to 100 participants prescreened in each category. Based on Prolific’s prescreening criterion only 26 participants in the oldest age category exhibit mental health disorders.
Significance
This feasibility study is exploring implicit objective methods of tracking mood and measuring depression severity which, if successful, will have important implications for diagnosis, evaluation of treatment efficacy and the development of novel targeted interventions. The proposed project measures mood swings in depression but does not require recalling or reflecting negative experiences explicitly; hence it may provide a comfortable way for patients to be assessed, potentially overcoming a practice concern in clinical settings.
Open science commitment
This study is pre-registered. Once the study is complete all study materials, analysis code and data will be openly available on the Open Science Framework repository. We intend to publish in an open access journal.
References
[1] Cherry EC. Some Experiments on the Recognition of Speech, with One and with Two Ears. Citation: The Journal of the Acoustical Society of America 1953;25:975. https://doi.org/10.1121/1.1907229.
[2] Madan CR, Spetch ML. Is the enhancement of memory due to reward driven by value or salience? Acta Psychologica 2012;139. https://doi.org/10.1016/j.actpsy.2011.12.010.
[3] Frings C, Wentura D. Self-Priorization Processes in Action and Perception Visuo-tactile Representational Momentum View project The social meaning of emotions View project. Article in Journal of Experimental Psychology Human Perception & Performance 2014. https://doi.org/10.1037/a0037376.
[4] Sui J. Perceptual Effects of Social Salience: Evidence From Self-Prioritization Effects on Perceptual Matching. Article in Journal of Experimental Psychology Human Perception & Performance 2012. https://doi.org/10.1037/a0029792.
[5] Sui J, Ohrling E, Humphreys GW. Negative mood disrupts self- and reward-biases in perceptual matching. Quarterly Journal of Experimental Psychology 2016;69:1438–48. https://doi.org/10.1080/17470218.2015.1122069.
[6] Stolte M, Humphreys G, Yankouskaya A, Sui J. Dissociating biases towards the self and positive emotion. Quarterly Journal of Experimental Psychology 2017;70:1011–22. https://doi.org/10.1080/17470218.2015.1101477.
[7] McIvor L, Sui J, Malhotra T, Drury D, Kumar S. Self-referential processing and emotion context insensitivity in major depressive disorder. European Journal of Neuroscience 2021;53. https://doi.org/10.1111/ejn.14782.
[8] Beck AT. Cognitive therapy and the emotional disorders. INTUNIVPRESS,NEW YORK 1976. https://doi.org/10.1176/appi.psychotherapy.1977.31.4.633.
[9] Komarahadi FL, Maurischat C, Härter M, Bengel J. Relationship of depression and anxiety with social desirability in chronic pain patients. Schmerz 2004;18. https://doi.org/10.1007/s00482-003-0282-2.
[10] Eshel N, Roiser JP. Reward and punishment processing in depression. Biological Psychiatry 2010;68. https://doi.org/10.1016/j.biopsych.2010.01.027.
[11] Julious SA. Sample size of 12 per group rule of thumb for a pilot study. Pharmaceutical Statistics 2005;4. https://doi.org/10.1002/pst.185.
[12] Sim J, Lewis M. The size of a pilot study for a clinical trial should be calculated in relation to considerations of precision and efficiency. Journal of Clinical Epidemiology 2012;65. https://doi.org/10.1016/j.jclinepi.2011.07.011.
[13] Anvari F, Lakens D. Using Anchor-Based Methods to Determine the Smallest Effect Size of Interest. Using Anchor-Based Methods to Determine the Smallest Effect Size of Interest 2019. https://doi.org/10.31234/osf.io/syp5a.
[14] Qian H, Wang Z, Li C, Gao X. Prioritised self-referential processing is modulated by emotional arousal. Quarterly Journal of Experimental Psychology 2020;73. https://doi.org/10.1177/1747021819892158.
[15] Watson D, Clark LA, Tellegen A. Development and validation of brief measures of positive and negative affect: The PANAS scales. Journal of Personality and Social Psychology 1988;54. https://doi.org/10.1037//0022-3514.54.6.1063.
[16] Beck AT, Steer RA, Brown GK. Manual for the Beck depression inventory-II. San Antonio, TX: Psychological Corporation 1996.
[17] Ryff CD, Keyes CLM. The Structure of Psychological Well-Being Revisited. Journal of Personality and Social Psychology 1995;69. https://doi.org/10.1037/0022-3514.69.4.719.
[18] Zeger SL, Karim MR. Generalized linear models with random effects; a Gibbs sampling approach. Journal of the American Statistical Association 1991;86. https://doi.org/10.1080/01621459.1991.10475006.
[19] Breslow NE, Clayton DG. Approximate Inference in Generalized Linear Mixed Models. Journal of the American Statistical Association 1993;88. https://doi.org/10.2307/2290687.