What you don’t know,can’t hurt you: The differential effect of masked versus non-masked counterconditioning and mere exposure to spider pictures on women’s affective evaluation of spiders

Abstract

Exposure therapy is the treatment of choice for specific phobias but prolonged exposure to feared stimuli is strenuous and may lead to treatment dropout. Previous research showed that repeated exposure to masked spiders was effective in reducing psychophysiological and behavioural fear responses, but appeared ineffective in changing subjective feelings towards spiders. This study investigated in an unselected female sample if masked counterconditioning would be more effective in reducing spider dislike compared to masked exposure, and if masked counterconditioning would also be more effective than non-masked counterconditioning. Women with varying levels of spider aversion (N = 272) were randomly assigned to one of four conditions. Three spider pictures were always (counterconditioning) or never (exposure) followed by smiling faces. For half of the participants in each condition the spiders were masked. Results indicated that participants rated the spider more positively after both masked counterconditioning and masked exposure. However, the increase in valence after masked counterconditioning was not significantly larger than after mere masked exposure, or after non-masked counterconditioning. Thus, our findings show that repeated exposure to masked spider pictures is effective in reducing spider aversion, but they provided no support for the anticipated added benefit of pairing the spider with positive stimuli.

Keywords

Continuous flash suppression evaluative counterconditioning exposure masking spider phobia

Introduction

Spider fear is very common in the general population, especially among women (Arrindell, 2000; Davey, 1994; Stinson et al., 2007; Wardenaar et al., 2017). Exposure therapy has been proven effective in treating spider fear (e.g. Arntz & Lavy, 1993; Öst et al., 1991), and is considered the treatment of choice for specific phobias. However, only a relatively low number of individuals seek treatment for their (spider) phobia (Stinson et al., 2007). The anticipated distress that is caused by the (prolonged) exposure to the feared stimulus might be one of the factors that reduces people’s willingness to engage in an exposure-based treatment.

Since limiting perceptual awareness of fear-eliciting stimuli during exposure may increase individuals’ willingness to enter and complete treatment, researchers have begun to study the impact of repeated exposure (EXP) to masked phobic stimuli. A series of studies by Siegel and colleagues (2009, 2011, 2012, 2018) showed that individuals with high levels of spider fear were more likely to approach a real tarantula after EXP to pictures of backwardly-masked spiders (i.e. very short spider presentations immediately followed by longer presentations of a masking array of letters) than after a control procedure with masked flowers or EXP to visible spider pictures. Masking reduced participants’ distress during the EXP procedure, and relatively low distress during masked EXP was associated with less subjective fear while approaching the real tarantula after EXP (Siegel et al., 2018). A review of 26 studies concluded that EXP to masked animal pictures does result in successful extinction of psychophysiological and behavioural fear responses (e.g. avoidance), however, it appeared to hardly improve the subjective feelings towards the feared animal (Frumento et al., 2021).

This lack of improvement in subjective feelings after (masked) EXP may increase the likelihood that psychophysiological and behavioural fear responses over time return, and it may be responsible for the rather high incidence of (partial) return of fear after EXP (incidence estimates range from 19 to 62%; Craske & Mystkowski, 2006). Systematically pairing a feared animal with inherently positive stimuli (i.e. evaluative counterconditioning; CC) might result in more long-lasting reductions in phobic behaviour than repeated animal-only presentations (EXP) because CC might be more effective in changing the subjective feelings towards the animal. Several studies showed that CC can lead to a reduction in the negative valence of stimuli whereas EXP does not (Baeyens et al., 1989; Engelhard et al., 2014; Kerkhof et al., 2011; van Dis et al., 2019) or only to a lesser extent (Eifert et al., 1988). Also, reinstatement of fear has been found less likely after CC than EXP (Kang et al., 2018; Keller & Dunsmoor, 2020; Kerkhof et al., 2011). Together these findings suggest that CC-based interventions might outperform EXP in the treatment of animal fears (in the long term).

Just like EXP procedures seem to benefit from masking feared stimuli (Siegel and Weinberger, 2009, Siegel et al., 2011, Siegel and Weinberger, 2012, Siegel et al., 2018), masked CC procedures may be less likely to cause distress and could lead to better outcomes than non-masked (common) CC. A fundamental rule of aversive-to-appetitive conditioning is that the positive response to the unconditional stimulus (US) should be stronger than the negative response to the conditional stimulus (CS; Keller et al., 2020). A benefit of masking feared animals (CS) during CC is that it may prevent the activation of strong negative affective responses which may otherwise counteract the impact of the (paired) positive US, thereby weakening or even undermining the effects of the CC procedure as a means to reduce the negative response to feared animals. Also, humans tend to give more weight to negative than positive events (Rozin & Royzman, 2001), and limiting the visibility of the feared animals during CC may prevent that positive USs are discarded.

Whether conditioning effects can be obtained with masked procedures is up for debate with some researchers arguing that masked EC is just as effective as non-masked EC (Hofmann et al., 2010), and others arguing that CS+ visibility is necessary to induce EC effects (Stahl et al., 2016). Crucially, none of these studies looked at a CC procedure with a priori feared stimuli, which may be more likely to be sufficiently processed when masked (e.g. Siegel et al., 2017). To our knowledge, there are only two earlier studies that tested a CC paradigm with invisible feared animals as CS (Homan et al., 2021; Taschereau-Dumouchel et al., 2018). Neither study assessed the subjective evaluation of the CS+ as an outcome measure, however, their results do support the idea that (physiological) affective responses to animals can be manipulated by a CC procedure during which the animals (CS) are not (clearly) visible.

The current study was designed to test if masked CC is indeed more effective than both masked EXP and non-masked CC in changing the valence of common house spiders. Given that we aimed to test an unproven and experimental procedure, we decided to first conduct a proof-of-principle study and tested our hypotheses in an unselected sample of female students. Spider fears are very common among students (Seim & Spates, 2009) and especially women (Arrindell, 2000). Our first goal was to investigate whether masked CC is more effective than masked EXP in reducing the negative valence of the spider. Our second goal was to compare the effectiveness of masked versus non-masked CC. In both the masked and non-masked CC conditions, the spider CS was always followed by a 600 milliseconds positive US (i.e. a smiling face), whereas USs were never presented after the spider CS in the masked and non-masked EXP conditions. Continuous flash suppression (Tsuchiya & Koch, 2005) was used to suppress the CS in the masked conditions (the USs were always visible). We predicted that the spider CS would be rated more positively after masked CC than after masked mere EXP (hypothesis 1) and non-masked CC (hypothesis 2). In addition to assessing the effect of masked CC on an initially negative CS (i.e. the spider), we also checked if our conditioning procedures could be used to change the valence of an initially neutral stimulus (i.e. a coffee cup). We predicted that the neutral CS would be rated more positively after it was systematically paired with positive USs than after repeated CS-only presentations (evaluative conditioning effect; Hofmann et al., 2010), both when the CS was masked (hypothesis 3) and when the CS was clearly visible (hypothesis 4).

Method

Participants

A total of 272 female participants (M age = 20.3, SD = 2.16) were recruited via the participant databases of the University of Groningen. They received money or course credits for their participation. Exclusion criteria were photosensitivity, colour blindness, and glasses, unless participants agreed to wear contact lenses to the lab or had normal near sight vision and could participate without their glasses (glasses did not fit under the mirror stereoscope). Prior to the statistical analyses, data of 13 participants were excluded because they self-indicated to not have filled out the questionnaires seriously (n = 1); because the CC or EXP procedure stopped working (n = 2); or because errors were made by the experimenter during the early phase of data collection (n = 10). The participants were randomly assigned to the masked CC condition (n = 62), masked EXP condition (n = 66), non-masked CC condition (n = 66), and non-masked EXP condition (n = 65). The present study was approved by the Ethical Committee Psychology of the University of Groningen under the code PSY-2122-S-0033. The study was preregistered https://aspredicted.org/6v45a.pdf.

Materials

Stimuli

Spider CSs were three pictures of a house spider (Tegenaria atrica) taken from the front, left, and right side. Two of these pictures were derived from the Internet (Groenewegen, 2012a, 2012b) and the third was obtained by mirroring the right side picture in Window’s picture editor. Neutral CSs were three pictures of a light-blue coffee cup that were taken from the left, right, and top with a Panasonic DC-FZ82 camera. We also selected 9 neutral pictures from the International Affective Picture System (Lang et al., 1997) to use as filler stimuli in the CC and EXP procedures. To facilitate masking, the settings of the CSs and filler stimuli were changed to −100 contrast and −100 light in Window’s picture editor. USs were 15 smiling faces (8 male/7 female models) from the NimStim Facial Stimuli Set (Tottenham et al., 2009). Masks were 100 different colourful patterns made in Matlab using an open-source code. To hide our interest in spider fear, we also selected 9 animal pictures (i.e. a pigeon, mouse, fly, cat, butterfly, dog, puppy, and rabbit) from the Internet to use as filler stimuli in the rating procedure (these were not used in the CC and EXP procedures). The animals included a second house spider (Steatoda grossa from the right side) to see if changes in the appreciation of the spider CSs generalized to this other spider picture. We removed the background of all pictures and cropped each stimulus to 352 (width) × 452 (length) pixels. The stimulus identification numbers of the neutral filler stimuli and USs can be found in supplementary file A.

Manipulation

The CC and EXP procedures were programmed using OpenSesame version 3.3.6 (Mathôt et al., 2012). There were two blocks of 60 trials of which were 30 neutral filler trials (total duration +/− 15 minutes). Block 1 consisted of 30 neutral coffee cup CS trials and Block 2 of 30 spider CS trials. All stimuli were presented for 4 seconds, except for the USs which were presented for 600 milliseconds. The inter-trial-interval had a variable duration (i.e. 1, 2, or 3 seconds) and during this time only two side-by-side black squares (427.6 × 527.6 pixels) were visible. Stimuli were in all conditions presented at the centre of the black square on the side of participants’ non-dominant eye. The black frame aided continued overlapping of simultaneously presented stimuli in the masked conditions. Within each block, the order of the stimuli (and masks, if applicable) was random. The version of the procedure that participants received depended on the condition that they were assigned to (see Figure 1).

Figure 1.

Example of a CS trial in block 1 (a) and block 2 (b) of the masked CC condition (1), masked EXP condition (2), non-masked CC condition (3), and non-masked EXP condition (4). Stimuli were always presented in the non-dominant eye. In the masked conditions, the percept consisted of the colourful patterns that were presented in the dominant eye.

The first block was included to test the efficacy of the (masked) conditioning procedure with regard to neutral stimuli, whereas the second block was included to test the differential effect of masked versus non-masked CC and EXP to spider stimuli. In the masked and non-masked CC conditions, the spider CS was always followed by a positive US and filler neutral stimuli were followed by a blank screen of the same duration as the USs (i.e. for 600 milliseconds only one or two black squares were visible). In these conditions, the neutral CS and filler stimuli in block 1 were both always followed by a blank screen (and thus never by a positive US). In the masked and non-masked EXP conditions, the spider CS was always followed by a blank screen (never a positive US) and the neutral CS in block 1 always by a positive US. In these conditions, the filler stimuli were always followed by a blank screen (cf. the CC procedures). This approach thus allowed us to examine the effects of (masked) CC and (masked) EXP on the affective evaluations of both spider stimuli and initially neutral stimuli while securing that in all conditions the exposure to positive USs was exactly similar.

In the masked conditions, 30 masks were presented simultaneously with the CS or filler stimulus in the dominant eye for 0.1 second each (i.e. they flashed at a frequency of 10 Hz). We chose this method of masking (i.e. CFS) because previous research indicated it is more reliable and allows for longer stimulus presentation times compared to other masking techniques (Tsuchiya & Koch, 2005), and the latter may promote contingency learning. Furthermore, we wanted to keep the masking technique and some other elements of the CC procedure (e.g. the USs) consistent with our previous work on masked CC in the context of body image (Masselman et al., 2023). No masks were presented during the presentation of the USs. The CC and EXP procedures (including masks and animal filler stimuli) can be obtained via https://doi.org/10.34894/QBINQO.

Measures

Affective ratings

The stimuli were presented one-by-one in Qualtrics and rated on valence using a Visual Analogue Scale (VAS) underneath it (0 = Very negative and 100 = Very positive). Before the manipulation, participants rated the neutral coffee cup CS, spider CS, second ‘generalization’ spider, other filler animals, USs, and filler neutral stimuli. After the first block, the neutral coffee cup CS, USs, and filler neutral stimuli were rated again (allowing to test the impact of the [masked] evaluative conditioning procedure on the neutral CS). After the second block, the spider CS, generalization spider, other filler animals, and USs were rated again. The order of the stimulus categories was as listed above, but within each category the stimuli were presented in a random order. The ratings of the USs were averaged. Higher ratings indicated a more positive evaluation of the stimulus/stimuli (range = 0–100).

Mental behavioural approach test

For exploratory purposes, a mental behavioural approach test (BAT) was used to assess participants’ (mental) willingness to approach the spider CS after the manipulation. Participants were instructed to imagine that the spider CS was placed in a glass jar on a table and that they had to approach it. Underneath the spider CS (frontal picture) they selected to what step they were willing to go: none (0); walk towards the spider as near as you can (1); touch the jar (2); open the jar (3); take the jar in your hands (4); touch the spider with a pencil (5); put the spider in a washing bowl (6); touch the spider with a finger (7); or let the spider walk over your hands (8; these steps were consistent with de Jong et al., 2000). The mental BAT score was the final step participants were willing to take, and higher scores thus indicated a stronger (mental) approach to the spider CS.

We planned to also assess participants’ subjective fear while imagining executing the final step of the mental BAT. Due to a technical error, the instruction (‘Imagine that this spider walks over your hands. How fearful would you be?’) was not displayed and we therefore had to omit this measure from our planned analyses.

Spider fear

The abbreviated Spider Phobia Questionnaire (SPQ-15; Olatunji et al., 2009) was used as a descriptive measure of participants’ fear and avoidance of spiders. It has 15 items that are answered with ‘True’ or ‘False’. The items answered with ‘True’ are summed to obtain a total score, and higher scores indicate stronger spider fear (range = 0–15). The SPQ-15 had a good internal reliability (Cronbach’s α = .86).

The Spider Distress Scale (SDS; Peters et al., 2022) was used as a second measure of spider fear and avoidance. Its 17 items are answered on a 7-point Likert scale (0 = Completely disagree and 6 = Completely agree). The SDS has a spider fear subscale (items 1–13; range = 0–78) and spider disgust subscale (items 14–17, range = 0–24). Total and subscale scores can be obtained by summing the relevant items. Higher SDS total scores indicate stronger spider fear (range = 0–102). The internal reliability was good for the SDS (Cronbach’s α = .95) and its subscales (Fear subscale: Cronbach’s α = .95; disgust subscale: Cronbach’s α = .81).

Procedure

This study was advertised as assessing the effect of seeing (dis)liked stimuli on mood. After participants gave written informed consent, we identified their dominant eye and asked them whether their vision was corrected. Participants then rated the valence of the stimuli. Prior to the manipulation, participants received verbal instructions to sit still and view the stimuli (non-masked conditions) or colours (masked conditions) on the computer screen (1920 × 1080 IIYAMA ProLite G2773HS, refresh rate = 100 Hz). Participants in the masked conditions viewed these through a mirror stereoscope with chinrest (ScreenScope Pro Mirror Stereoscope, Forestry Suppliers Inc) which they first had to calibrate. We made sure our instructions concealed the function of the mirror stereoscope (e.g. by avoiding terms such as ‘masking’ and ‘overlapping stimuli’). After the first and second block, participants rated the valence of the presented stimuli. Finally, they completed the mental BAT, question about their fear of the spider CS, SPQ-15, SDS, age, and questions about demand awareness, contingency awareness, and, in the masked conditions, stimulus awareness (in this order). The different types of awareness were assessed in a stepwise manner (see supplementary file B). This was done at the end of the test session to prevent drawing attention to the suppressed stimuli during the manipulation.

Statistical analyses

Consistent with the preregistration (https://aspredicted.org/6v45a.pdf), four analyses of covariance (ANCOVAs) were performed with post-manipulation CS valence as the dependent variable, condition as the independent variable, and pre-manipulation CS valence as the covariate. The valence of the spider CS was compared between the masked CC condition and masked EXP condition (hypothesis 1) as well as between the masked CC condition and the non-masked CC condition (hypothesis 2) in two separate ANCOVAs. The valence of the neutral CS was compared between participants for whom the neutral CS was always followed by positive USs and participants for whom the neutral CS was never followed by USs. This was done for the masked procedures (hypothesis 3) and non-masked procedures (hypothesis 4) separately. The primary analyses were repeated without: outliers (i.e. data points that exceeded 1.5 times the interquartile range); participants in the masked conditions who report to have seen the spider CS; participants who correctly estimated the CS-US contingency; and participants who were demand aware (i.e. indicated that the manipulation was meant to improve their appreciation of the CS). Alpha criterion was set to .0125 (α = .05/4) for the primary analyses to correct for multiple testing. For exploratory purposes, we also compared the listed conditions on post-manipulation mental BAT scores (ANOVA) as well as valence ratings of the generalization spider and USs (ANCOVAs). Exploratory analyses assessing the i) effect of masked versus non-masked EXP on the spider CS; ii) effect of masked versus non-masked conditioning on the neutral CS; iii) interaction effect of condition and spider fear levels (i.e. SPQ scores); and iv) main effects of masking and procedure, can be found in the supplementary file C. For the exploratory analyses, an uncorrected alpha criterion of .05 was used.

Power calculation

A priori power analyses using G*Power 3.1 (Faul et al., 2009) indicated that a sample of 182 participants provides a power of .80 to detect a medium effect (f = .25) at an alpha level of .0125 (α = .05/4). To make sure we had sufficient power, we preregistered (https://aspredicted.org/6v45a.pdf) that we wanted to test 256 participants (i.e. approximately 64 participants per condition) and ended up with a sample of 259 participants.

Results

Participants

For an overview of the condition means, standard deviations, and post hoc non-parametric Kruskal–Wallis tests per descriptive variable, see Table 1. Most participants had normal uncorrected vision (n = 158), wore contact lenses (n = 59), or could participate without glasses/lenses because they had normal near sight vision (n = 39). One participant was allowed to participate with glasses because she was already randomly assigned to a non-masked condition and two participants reported to have (corrected) amblyopia. They were not excluded because during the calibration of the mirror stereoscope they appeared to be sensitive to binocular suppression. A total of 165 participants showed dominance of the right eye and 94 participants of the left eye. Although the vast majority of participants (n = 253) rated the spiders as negative (i.e. below 50), the overall SPQ score (M = 5.92; SD = 3.69) was intermediate between previously reported mean scores of phobic and non-phobic individuals (e.g. Spider phobia: M = 10.7, SD = 1.79; no spider phobia: M = 0.86, SD = 2.04; Olatunji et al., 2009).

Table 1.

Condition means, standard deviations, and post hoc Kruskal–Wallis tests per descriptive variable.

		Masked (n = 128)		Non-masked (n = 131)		H(3)	p
		CC (n = 62)	EXP (n = 66)	CC (n = 66)	EXP (n = 65)
		M (SD)	M (SD)	M (SD)	M (SD)
Descriptives variables	Age	20.06 (1.99)	20.35 (2.09)	20.20 (2.46)	20.58 (2.21)	3.47	.33
	SPQ-15	6.11 (3.54)	5.62 (3.54)	6.39 (4.01)	5.55 (3.64)	2.21	.53
	SDS	53.52 (21.85)	50.05 (23.16)	56.36 (24.81)	51.28 (22.47)	3.57	.31
	Fear	38.37 (18.49)	35.76 (18.71)	40.79 (19.95)	36.55 (18.06)	3.33	.34
	Disgust	15.15 (5.50)	14.29 (6.03)	15.58 (6.23)	14.72 (5.56)	2.03	.57

Note. SPQ-15 = Abbreviated Spider Phobia Questionnaire (range 0–15, higher scores indicate stronger spider fear); SDS = Spider Distress Scale (range 0–102, higher scores indicate stronger spider fear); fear = fear subscale of the Spider Distress Scale (range 0–78, higher scores indicate stronger spider fear); disgust = disgust subscale of the Spider Distress Scale (range 0–24, higher scores indicate stronger spider disgust).

Primary analyses

Affective ratings spider CS

For an overview of the means and standard deviations of the outcome variables pre- and post-manipulation per condition, see Table 2. Analyses indicated no violations of the assumption of normality and linearity.¹ The variance and regression slope of the masked CC condition appeared homogeneous with that of the masked EXP condition (Variance: F(1, 126) = 0.78, p = .38; slope: F(1, 124) = 1.60, p = .21) and non-masked CC condition (Variance: F(1, 126) = 0.22, p = .64; slope: F(1, 124) = 0.12, p = .73).

Table 2.

Means and standard deviations for outcome variables pre- and post-manipulation per condition.

		CC		EXP
		Pre-test	Post-test	Pre-test	Post-test
		M (SD)	M (SD)	M (SD)	M (SD)
Masked (n = 128)	Participants	62	62	66	66
	Valence
	Spider CS	15.84 (15.09)	21.85 (19.25)	15.52 (16.78)	21.48 (16.72)
	Spider 2	16.13 (15.56)	19.26 (16.72)	16.74 (15.87)	20.45 (17.95)
	USs	71.12 (12.81)	68.45 (13.52)	69.92 (12.97)	68.60 (13.39)
	Mental BAT	—	3.19 (2.39)	—	4.12 (2.31)
Non-masked (n = 131)	Participants	66	66	65	65
	Valence
	Spider CS	15.59 (15.96)	17.64 (20.14)	20.72 (16.72)	22.15 (19.52)
	Spider 2	16.20 (17.18)	16.00 (19.98)	20.55 (16.32)	20.54 (17.79)
	USs	70.04 (13.12)	65.65 (14.30)	70.98 (11.78)	69.61 (12.38)
	Mental BAT	—	3.29 (2.33)	—	3.88 (2.14)

Note. Participants = number of participants; valence (range 0–100; higher scores indicate a more positive evaluation of the CS); spider 2 = valence of generalization spider that was not part of the CC and EXP procedure; USs = mean valence of the USs after the second block of the manipulation; mental BAT (range 0–9; higher scores indicate a stronger approach to the spider CS); subjective fear (range 0–100; higher scores indicate more fear of the spider CS); valence USs (range 0–100; higher scores indicate a more positive evaluation of the USs).

Masked CC versus masked EXP

The results from the first ANCOVA indicated that pre-manipulation valence significantly predicted post-manipulation valence of the spider CS (F(1, 125) = 123.2, p < .001, η_p² = .50) and that, contrary to our first hypothesis, post-manipulation valence ratings of the spider CS in the masked CC condition did not differ significantly from the ratings in the masked EXP condition (F(1, 125) = 0.003, p = .96, η_p² < .001). For a more comprehensive appreciation of the results, we post hoc explored the change in valence within each condition. Two paired sample t-tests indicated that there was a significant increase in the valence of the spider CS in the masked CC condition (t(61) = −3.44, p < .001, d = .44) and the masked EXP condition (t(65) = −3.85, p < .001, d = .48). Together, these findings indicate that the spider CS valence ratings increased after both masked CC and masked EXP but that the improvement in valence was not larger for the masked CC condition than for the masked EXP condition.

Masked CC versus non-masked CC

Results from the second ANCOVA indicated that pre-manipulation valence ratings significantly predicted post-manipulation ratings of the spider CS in the CC conditions (F(1, 125) = 142.9, p < .001, η_p² = .53) but, contrary to our second hypothesis, there was no significant difference in post-manipulation spider CS valence ratings between the masked CC and non-masked CC conditions (F(1, 125) = 2.78, p = .10, η_p² = .02). There was a non-significant trend in the predicted direction. A post hoc paired sample t-test indicated that there was no significant pre- to post-manipulation increase in the valence of the spider CS in the non-masked CC condition (t(65) = −1.25, p = .22, d = .16). Thus, there was no significant increase in the valence of the spider CS after the non-masked CC procedure but contrary to our second hypothesis, the difference in post-intervention spider CS ratings between the non-masked and masked CC condition did not reach statistical significance.

Awareness check

The pattern of the results from the primary analyses did not change when we performed the ANCOVAs without contingency aware participants (Masked CC condition: n = 7; masked EXP condition: n = 1; non-masked CC condition: n = 4; non-masked EXP condition: n = 1), demand aware participants (Masked CC condition: n = 6; masked EXP condition: n = 3; non-masked CC condition: n = 4; non-masked EXP condition: n = 11), or participants in the masked conditions who indicated to have (incidentally) seen the spider CS (Masked CC condition: n = 36; masked EXP condition: n = 36).²

Affective ratings neutral CS

Due to apparent heterogeneity of the variance in the masked conditions (F(1, 126) = 3.82, p = .05) and of the regression slopes in the non-masked conditions (F(1, 127) = 7.73, p = .01), non-parametric rank ANCOVAs (Quade, 1967) were performed.

Masked EC versus masked CS-only presentations

The results from the first non-parametric rank ANCOVA indicated that systematically pairing the masked neutral CS with positive USs (EC) did not improve the evaluation of the neutral CS more than a control procedure with masked CS-only presentations (F(1, 126) = 1.62, p = .21, η² = .01), which was in contrast to our third hypothesis. To better understand the pattern of findings, we post hoc explored the within-subjects effects of the masked EC and masked CS-only procedure using non-parametric Wilcoxon matched-pair signed-rank tests. The results indicated a significant decrease in valence of the neutral CS after masked CS-only presentations (Pre: M = 64.2, SD = 15.7; post: M = 59.8, SD = 15.8; t(61) = −2.77, p = .01, d = .35) but there was no significant increase/decrease in neutral CS valence after masked EC (Pre: M = 58.9, SD = 14.5; post: M = 59.1, SD = 12.8; t(65) = −0.42, p = .68, d = .08). Thus, we observed a decrease in neutral CS valence in the masked CS-only condition that was absent in the masked EC condition, but there was no significant difference in post-manipulation valence ratings of the neutral CS between these conditions.

Non-masked EC versus non-masked CS-only presentations

A second Quade test indicated that, contrary to our fourth hypothesis, there was no difference in post-manipulation valence ratings of the neutral CS after non-masked EC versus non-masked CS-only presentations (F(1, 129) = 2.28, p = .13, η² = .02). Additional exploratory non-parametric Wilcoxon matched-pair signed-rank tests indicated a significant pre- to post-manipulation decrease in neutral CS valence ratings in both the non-masked EC condition (Pre: M = 62.7, SD = 14.1; post: M = 57.0, SD = 13.6; t(64) = −2.85, p = .004, d = .35) and the non-masked CS-only presentations condition (Pre: M = 61.4, SD = 17.3; post: M = 51.8, SD = 22.7; t(65) = −4,06, p < .001, d = .50). Pairing non-masked neutral CSs with positive USs thus did not increase the valence of the neutral CS more than a control procedure during which the non-masked CSs were never paired with positive USs (instead we observed a decrease in valence over time in both conditions).

Secondary analyses

Mental BAT

Results indicated that the differences between the conditions in post-manipulation mental BAT scores approached significance (F(3, 255) = 2.50, p = .06, η² = .03). Post hoc LSD tests indicated that participants in the masked EXP condition scored higher on the mental BAT (i.e. showed more mental approach to the spider CS) than participants in the masked CC condition (M _difference = 0.93, p = .02) and participants in the non-masked CC condition (M _difference = 0.83, p = .04). For the comparison of the masked and non-masked EXP condition, see supplementary file C.

Affective ratings generalization spider

Participants in the masked CC condition gave the generalization spider more positive ratings after the manipulation than participants in the non-masked CC condition (rank ANCOVA: F(1, 126) = 4.14, p = .04, η² = .03). There was no difference in post-manipulation valence ratings of the generalization spider between the masked CC and masked EXP conditions (ANCOVA: F(1, 125) = 0.11, p = .74, η² = .001). Post hoc paired samples t-tests indicated significant pre- to post-manipulation increases with regard to the valence ratings of the generalization spider in the masked CC condition (t(61) = −2.13, p = .04, d = .27) and masked EXP condition (t(64) = −2.45, p = .02, d = .30), but not for the non-masked CC condition (t(65) = 0.13, p = .90, d = .02) or non-masked EXP condition (see supplementary file C).

Affective ratings USs

Valence ratings prior to the manipulation indicated that the USs on average were indeed considered positive (M = 70.5, SD = 12.6) and the filler stimuli indeed neutral (M = 48.4, SD = 7.3). Participants in the non-masked CC condition (i.e. the condition in which the USs were always preceded by visible spiders) rated the USs more negatively after the manipulation than participants in the non-masked EXP condition (for whom the USs were never paired with spiders; ANCOVA: F(1, 128) = 4.85, p = .03, η_p² = .04). There was no significant difference in post-manipulation valence of the USs between the masked CC condition and masked EXP condition (rank ANCOVA: F(1, 126) = 0.61, p = .44, η² = .005), or masked CC condition and non-masked CC condition (ANCOVA: F(1, 125) = 1.40, p = .24, η_p² = .01).³

Discussion

The present study tested if a counterconditioning (CC) procedure in which masked pictures of a spider (CS) were always followed by a smiling face (US) was more effective in reducing the negative valence of the spider than a masked spider-only exposure (EXP) procedure, or a non-masked CC procedure in which the spider CS was clearly visible. The results indicated that spider CS ratings were not more positive (or negative) after masked CC than after masked EXP (≠ hypothesis 1), or non-masked CC (≠ hypothesis 2). The spider CS on average became more positive after both masked CC and masked EXP, whereas the valence ratings of the spider CS did not significantly change after non-masked CC (or non-masked EXP). We also examined the effect of these procedures with a neutral coffee cup as CS and found that the neutral CS was not rated more positively after a procedure in which it was systematically followed by positive USs than after repeated CS-only presentations. This was independent of whether the CS was masked during the conditioning and control procedure (≠ hypothesis 3) or not (≠ hypothesis 4).

In contrast to research suggesting that masked EXP does not affect subjective feelings towards feared animals (Frumento et al., 2021), our findings indicate that simply exposing individuals to a masked spider was sufficient to improve the evaluation of not only this spider but also generalized to a second spider that was never part of the EXP procedure. There was no added benefit to pairing the masked spider with positive USs, which is inconsistent with previous studies showing that CC had a larger impact on the negative valence of stimuli than CS-only extinction procedures (Baeyens et al., 1989; Eifert et al., 1988; Engelhard et al., 2014; Kerkhof et al., 2011; van Dis et al., 2019). It is, however, consistent with studies that failed to demonstrate a superior effect of CC procedures on CS valence compared to extinction procedures (de Jong et al., 2000; Gatzounis et al., 2022; Raes & De Raedt, 2012). Since we did not include a mere repeated assessment control condition (i.e. without presenting spider stimuli between pre-post assessments of spider valence), it cannot be ruled out that the improvement in spider valence ratings after masked EXP and CC was driven by something other than the manipulation, for instance a regression to the mean (i.e. less extreme scores during the second assessment). However, when we compared the pre- and post-manipulation valence ratings of the filler animals (not present during the manipulation), we did not find evidence supporting this alternative explanation. That is, over time there was no significant increase in the valence of the (only) filler animal that was rated as negative at baseline (i.e. a housefly; pre-test: M = 28.1, SD = 14.7; post-test: M = 28.8, SD = 15.5). This, together with the stability of CS valence from pre- to post-manipulation in the non-masked conditions, indicates that the increase in positive valence of the spider CS and generalization spider over time does not reflect a general pattern of less extreme scores on the second assessment.

It is difficult to explain why we did not find a positive effect of counterconditioning on top of the improvement due to repeated exposure to the masked spider per se. We did find indirect evidence indicating that participants learnt the contingency between the spider CS and the smiling faces (USs) after non-masked (common) CC. The smiling faces were rated more negatively after they were systematically paired with visible spiders (non-masked CC condition) compared to when they were never paired with visible spiders (non-masked EXP condition). Such a drop in the appreciation of the smiling faces was not observed after the masked CC procedure. This is consistent with our starting point that masking would prevent the spider pictures from eliciting strong negative affective responses during the CC procedure. Yet, it did not have the expected positive effect on the efficacy of CC. Another explanation for the lack of valence transfer from the USs to the CS (or vice versa) could of course be that the masking prevented participants from learning the CS-US contingency.

The results with regard to the ‘neutral’ CS do not help clarify if participants are in fact susceptible for picking up CS-US contingencies when the CS is being masked. The coffee cup (CS) became less positive after all procedures except after the masked evaluative conditioning (EC) procedure. This could mean that there was a small effect of pairing the US with a masked CS that countered the general decline in valence over time, albeit that it is then hard to explain why common (non-masked) EC did not lead to a similar effect within the context of an a priori neutral stimulus (i.e. a stimulus that does not elicit a negative valence that can counteract the effectiveness of the EC procedure). Although CS-US belongingness does not seem to be a prerequisite for EC (Hofmann et al., 2010), participants may have learnt to associate the coffee cup (CS) with the visible filler items (other household items) instead of the smiling faces (US) because of their belongingness. Some ‘neutral’ filler items were on average rated as negative (e.g. a trash bin), which could explain the decrease in valence of the coffee cup over time. In the masked EC procedure the filler stimuli were masked, and the visibility of the smiling faces may have increased their saliency and ensured that participants learnt the contingency between the neutral CS and USs. It is, however, important to also keep in mind that the mean valence of the coffee cup was a little less positive at baseline in the masked EC condition, which may have influenced the results as well. Additional research is necessary before a more final conclusion can be made about the efficacy of masked conditioning procedures in changing affective appraisals of aversive stimuli.

Our findings conceptually replicate prior research showing that masked EXP can be applied to reduce spider aversion (Siegel and Weinberger, 2009; Siegel et al., 2011; Siegel and Weinberger, 2012; Siegel et al., 2018). Although some of these studies found better outcomes for masked compared to non-masked EXP, there was in the present study no significant difference in how the valence ratings of the spider CS or mental BAT scores changed after masked EXP compared to non-masked EXP (see the exploratory analyses in supplementary file C). One explanation for why there was no clear distinction between the effects of masked and non-masked EXP is that our analyses may have only been sensitive to picking up within-subjects changes but lacked the power to detect small-to-medium between-subjects effects. The contrast between the EXP conditions may have also been somewhat reduced by participants for whom the masking was suboptimal, albeit that the pattern of the results did not change when the analyses were repeated without the participants in the masked EXP condition who reported to have (occasionally) seen the CS (n = 36, 54.5%). It is important to note that our measure of CS+ awareness was quite conservative (e.g. participants were excluded even when there was an indication that they listed stimuli seen during the rating procedure and not the manipulation). Another reason for the relatively high level of participants who reported to have seen CSs+ is that there is some preliminary evidence suggesting large interindividual differences in the susceptibility to CFS (Gayet & Stein, 2017; Valuch & Mattler, 2019; Yamashiro et al., 2014), and participants may have started paying attention to the masked stimuli after becoming aware of their presence (e.g. accidentally after blinking). The finding that, compared to the relevant spider-only procedure, the valence of the USs became more negative after visible spider-US pairings and not after masked spider-US pairings suggests that overall CFS performance was sufficient to reduce at least some of the undermining effects of the aversive nature of the spider on the CC procedure. Another explanation for the comparable performance of masked and non-masked EXP is that our participants did not represent a sample of extremely fearful individuals. The contrast between the masked and non-masked EXP condition may have been reduced by participants who considered the spider negative/aversive but not frightening, and for whom it therefore did not matter whether the spider was masked or not (this also applies to the CC conditions), although we did not find a larger difference between the masked and non-masked CC or EXP conditions for participants with relatively high levels of spider fear (see supplementary file C). On a positive note, our findings indicate that masked EXP is not only beneficial for individuals with a spider phobia but that it might also reduce spider aversion in the non-clinical range. Since masked EXP is less strenuous for spider fearful individuals and may increase treatment compliance, it is promising that our results support that masked EXP is at least as effective as non-masked EXP in improving women’s subjective feelings towards spiders.

Whether stimuli contingencies can be learnt during masked conditioning procedures is an ongoing debate (e.g. Field, 2000). The efficacy of masked procedures may critically depend on the type of CSs that are involved. Accordingly, there is evidence that fear-relevant stimuli may be more likely to be processed while being masked compared to fear-irrelevant stimuli (Öhman et al., 1995), and masking may even aid the neural processing of fear-relevant stimuli (e.g. by preventing the activation of deficient fear regulatory processes; Siegel et al., 2017). The present study is the first to test the effectiveness of masked CC in changing subjective feelings towards fear-relevant stimuli, and it represents the first systematic comparison of masked versus non-masked conditioning and exposure procedures with aversive versus (ostensibly) neutral CSs. Our findings provided no support for the idea that masked CC can be applied to further reduce spider aversion (above and beyond what can be accomplished with masked EXP), and we found no conclusive evidence for masked conditioning as a means to change affective responses to fear-relevant or neutral stimuli. However, our failure to find an effect of a common (non-masked) EC procedure on the valence of a coffee cup might indicate that there were methodological shortcomings that need to be addressed before a conclusion can be drawn about the viability of masked conditioning procedures/interventions. An important aspect to consider, is that there is mounting evidence that affectively salient stimuli relatively easily break suppression by CFS (Gayet et al., 2016; Schmack et al., 2016). Given the great results that Siegel and colleagues (2009, 2011, 2012, 2018) achieved with their ‘very brief exposure’ technique (i.e. low levels of self-reported awareness of masked spiders despite participants’ high levels of spider fear), a logical next step would be to assess the efficacy of masked CC using a similar backward-masking technique as used in the studies of Siegel and colleagues. Individuals’ subjective experience of phobic stimuli should be an important target for intervention, and it is advised that future studies include subjective measures to adequately assess the efficacy of new interventions that target (animal) phobias. More research is necessary to extend our knowledge on applying evaluative (un)learning principles in treatment settings and to develop (spider)phobia treatments that are effective but lack the distress that is typically associated with the treatment of phobias.

Conclusions

The goal of the present study was to systematically compare the effects of masked and non-masked conditioning and exposure procedures on the evaluation of a spider and, for exploratory purposes, an ostensibly neutral coffee cup. Contrary to our hypotheses, the spider CS was not rated more positively after masked counterconditioning than after masked mere exposure or non-masked counterconditioning. When we looked at the within-subject effects, we found significant increases in the valence of the spider CS after masked counterconditioning and masked exposure but not after non-masked counterconditioning. The coffee cup was not rated more positively after masked and non-masked evaluative conditioning compared to a masked or non-masked control procedure. In line with previous research, our findings support that exposure to a masked spider is effective in reducing spider aversion. Although there was no added benefit to pairing the masked spider with positive stimuli, several limitations need to be addressed before a more final conclusion can be made about the efficacy of masked (counter)conditioning procedures.

Supplemental Material

Supplemental Material - What you don’t know, can’t hurt you: The differential effect of masked versus non-masked counterconditioning and mere exposure to spider pictures on women’s affective evaluation of spiders

Supplemental Material for What you don’t know, can’t hurt you: The differential effect of masked versus non-masked counterconditioning and mere exposure to spider pictures on women’s affective evaluation of spiders by Irina Masselman, Klaske A. Glashouwer and Peter J. de Jong in Journal of Experimental Psychopathology

Supplemental Material

Supplemental Material - What you don’t know, can’t hurt you: The differential effect of masked versus non-masked counterconditioning and mere exposure to spider pictures on women’s affective evaluation of spiders

Footnotes

Acknowledgments

The authors would like to thank Peter Groenewegen for the permission to use his artwork; Bert Hoekzema, Mark Span, and Pieter Zandbergen for their (technical) support; Mylène Nijp, Karien van Wieren, and Zoë Wogram for their help in collecting the data; and finally the students who volunteered to participate in this study.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported by an Idea Generator Grant [NWA.1228.191.406] that was awarded to the authors by the Netherlands Organization for Scientific Research. The funding source was not involved in the design of the study nor in the production of this manuscript.

CRediT authorship contribution statement

Irina Masselman: Conceptualization, Methodology, Software, Investigation, Formal analysis, Writing – Original Draft, Project administration, Funding acquisition. Klaske A. Glashouwer: Conceptualization, Methodology, Resources, Writing – Review and Editing, Supervision, Funding acquisition. Peter J. de Jong: Conceptualization, Methodology, Writing – Review and Editing, Supervision, Funding acquisition.

Pre-registration

The present study is preregistered at .

ORCID iDs

Irina Masselman

Klaske A. Glashouwer

Peter J. de Jong

Data availability statement

The raw data and SPSS syntax supporting the conclusions of this manuscript can be obtained via .

Supplemental Material

Supplemental material for this article is available online.

Notes

Author Biographies

Irina Masselman is a PhD candidate at the department of Clinical Psychology and Experimental Psychopathology at the University of Groningen (The Netherlands). She is currently working as a psychologist at Molemann Mental Health, a centre for specialized mental healthcare. Her research focusses on clinical applications for evaluative counterconditioning.

Klaske A. Glashouwer, PhD is a postdoctoral researcher at the department of Clinical Psychology and Experimental Psychopathology at the University of Groningen, and works as a trained cognitive behavioral therapist and researcher at the department of Eating Disorders of Accare, a centre for child and adolescent psychiatry in Groningen. Her research focuses on eating disorders.

Peter J. de Jong is full professor of Experimental Psychopathology at the University of Groningen. By combining experimental lab-based designs with (pre)clinical intervention studies his research focusses on delineating transdiagnostic mechanisms that contribute to the development and persistence of various disorders including anxiety disorders.

References

Arntz

Lavy

(1993). Does stimulus elaboration potentiate exposure in vivo treatment? Two forms of one-session treatment of spider phobia. Behavioural and Cognitive Psychotherapy, 21(1), 1–12. https://doi.org/10.1017/S0141347300017754

Arrindell

W. A.

(2000). Phobic dimensions: IV. the structure of animal fears. Behaviour Research and Therapy, 38(5), 509–530. https://doi.org/10.1016/S0005-7967(99)00097-2

Baeyens

Eelen

Van den Bergh

Crombez

(1989). Acquired affective- evaluative value: Conservative but not unchangeable. Behaviour Research and Therapy, 27(3), 279–287. https://doi.org/10.1016/0005-7967(89)90047-8

Craske

M. G.

Mystkowski

J. L.

(2006) Exposure therapy and extinction: Clinical studies. In Craske

M. G.

Hermans

Vansteenwegen

(Eds.), Fear and learning: From basic processes to clinical implications (pp. 217–233), American Psychological Association. https://doi.org/10.1037/11474-011

Davey

G. C.

(1994). Self-reported fears to common indigenous animals in an adult UK population: The role of disgust sensitivity. British Journal of Psychology, 85(4), 541–554. https://doi.org/10.1111/j.2044-8295.1994.tb02540.x

de Jong

P. J.

Vorage

van den Hout

M. A.

(2000). Counterconditioning in the treatment of spider phobia: Effects on disgust, fear and valence. Behaviour Research and Therapy, 38(11), 1055–1069. https://doi.org/10.1016/S0005-7967(99)00135-7

Eifert

G. H.

Craill

Carey

O’Connor

(1988). Affect modification through evaluative conditioning with music. Behaviour Research and Therapy, 26(4), 321–330. https://doi.org/10.1016/0005-7967(88)90084-8

Engelhard

I. M.

Leer

Lange

Olatunji

B. O.

(2014). Shaking that icky feeling: Effects of extinction and counterconditioning on disgust-related evaluative learning. Behavior Therapy, 45(5), 708–719. https://doi.org/10.1016/j.beth.2014.04.003

Faul

Erdfelder

Buchner

Lang

(2009). Statistical power analyses using G* power 3.1: Tests for correlation and regression analyses. Behavior Research Methods, 41(4), 1149–1160. https://doi.org/10.3758/BRM.41.4.1149

10.

Field

A. P.

(2000). I like it, but I’m not sure why: Can evaluative conditioning occur without conscious awareness? Consciousness and Cognition, 9(1), 13–36. https://doi.org/10.1006/ccog.1999.0402

11.

Frumento

Menicucci

Hitchcott

P. K.

Zaccaro

Gemignani

(2021). Systematic review of studies on subliminal exposure to phobic stimuli: Integrating therapeutic models for specific phobias. Frontiers in Neuroscience, 15(1). https://doi.org/10.3389/fnins.2021.654170

12.

Gatzounis

De Bruyn

Van de Velde

Meulders

(2022). No differences in return of pain-related fear after extinction and counterconditioning. Emotion, 22(8), 1886–1894. https://doi.org/10.1037/emo0000960

13.

Gayet

Paffen

C. L.

Belopolsky

A. V.

Theeuwes

Van der Stigchel

(2016). Visual input signaling threat gains preferential access to awareness in a breaking continuous flash suppression paradigm. Cognition, 149(1), 77–83. https://doi.org/10.1016/j.cognition.2016.01.009

14.

Gayet

Stein

(2017). Between-subject variability in the breaking continuous flash suppression paradigm: Potential causes, consequences, and solutions. Frontiers in Psychology, 8(1). https://doi.org/10.3389/fpsyg.2017.00437

15.

Groenewegen

(2012a). Gewone huisspin (tegenaria atrica) 6-2012 9019. [Photograph]. https://www.degroeneman.nl/fotografie/spinnen/gewone+huisspin+%28tegenaria+atrica%29+6-2012+9019.jpg.php

16.

Groenewegen

(2012b). Gewone huisspin (tegenaria atrica) 6-2012 9024. [Photograph]. https://www.degroeneman.nl/fotografie/spinnen/gewone+huisspin+%28tegenaria+atrica%29+6-2012+9024.jpg.php

17.

Hofmann

De Houwer

Perugini

Baeyens

Crombez

(2010). Evaluative conditioning in humans: A meta-analysis. Psychological Bulletin, 136(3), 390–421. https://doi.org/10.1037/a0018916

18.

Homan

Lau

H. L.

Levy

Raio

C. M.

Bach

D. R.

Carmel

Schiller

(2021). Evidence for a minimal role of stimulus awareness in reversal of threat learning. Learning & Memory, 28(3), 95–103. https://doi.org/10.1101/lm.050997.119

19.

Kang

Vervliet

Engelhard

I. M.

van Dis

E. A.

Hagenaars

M. A.

(2018). Reduced return of threat expectancy after counterconditioning versus extinction. Behaviour Research and Therapy, 108(1), 78–84. https://doi.org/10.1016/j.brat.2018.06.009

20.

Keller

N. E.

Dunsmoor

J. E.

(2020). The effects of aversive-to-appetitive counterconditioning on implicit and explicit fear memory. Learning & Memory, 27(1), 12–19. https://doi.org/10.1101/lm.050740.119

21.

Keller

N. E.

Hennings

A. C.

Dunsmoor

J. E.

(2020). Behavioral and neural processes in counterconditioning: Past and future directions. Behaviour Research and Therapy, 125(1). https://doi.org/10.1016/j.brat.2019.103532

22.

Kerkhof

Vansteenwegen

Baeyens

Hermans

(2011). Counterconditioning. Experimental Psychology, 58(1), 31–38. https://doi.org/10.1027/1618-3169/a000063

23.

Lang

P. J.

Bradley

M. M.

Cuthbert

B. N.

(1997). International affective picture system (IAPS): Technical manual and affective ratings. NIMH Center for the Study of Emotion and Attention, 1(1), 39–58. https://www2.unifesp.br/dpsicobio/adap/instructions.pdf.

24.

Masselman

Glashouwer

K. A.

Span

M. M.

de Jong

P. J.

(2023). The effectiveness of a masked counterconditioning approach using continuous flash suppression to alleviate body dissatisfaction in women with high body image concerns. Journal of Behavior Therapy and Experimental Psychiatry, 83(1). https://doi.org/10.1016/j.jbtep.2023.101938

25.

Mathôt

Schreij

Theeuwes

(2012). OpenSesame: An open-source, graphical experiment builder for the social sciences. Behavior Research Methods, 44(2), 314–324. https://doi.org/10.3758/s13428-011-0168-7

26.

Öhman

Esteves

Soares

J. J.

(1995). Preparedness and preattentive associative learning: Electrodermal conditioning to masked stimuli. Journal of Psychophysiology, 9(2), 99–108. https://psycnet.apa.org/record/1996-19315-001

27.

Olatunji

B. O.

Woods

C. M.

de Jong

P. J.

Teachman

B. A.

Sawchuk

C. N.

David

(2009). Development and initial validation of an abbreviated spider phobia questionnaire using item response theory. Behavior Therapy, 40(2), 114–130. https://doi.org/10.1016/j.beth.2008.04.002

28.

Öst

L. G.

Salkovskis

P. M.

Hellström

(1991). One-session therapist-directed exposure vs. self-exposure in the treatment of spider phobia. Behavior Therapy, 22(3), 407–422. https://doi.org/10.1016/S0005-7894(05)80374-0

29.

Peters

Visser

R. M.

Kindt

(2022). More than just fear: Development and psychometric evaluation of the Spider Distress Scale to assess spider fear and spider-related disgust. Journal of Anxiety Disorders, 90(1). https://doi.org/10.1016/j.janxdis.2022.102602

30.

Quade

(1967). Rank analysis of covariance. Journal of the American Statistical Association, 62(320), 1187–1200. https://doi.org/10.1080/01621459.1967.10500925

31.

Raes

A. K.

De Raedt

(2012). The effect of counterconditioning on evaluative responses and harm expectancy in a fear conditioning paradigm. Behavior Therapy, 43(4), 757–767. https://doi.org/10.1016/j.beth.2012.03.012

32.

Rozin

Royzman

E. B.

(2001). Negativity bias, negativity dominance, and contagion. Personality and Social Psychology Review, 5(4), 296–320. https://doi.org/10.1207/S15327957PSPR0504_2

33.

Schmack

Burk

Haynes

Sterzer

(2016). Predicting subjective affective salience from cortical responses to invisible object stimuli. Cerebral Cortex, 26(8), 3453–3460. https://doi.org/10.1093/cercor/bhv174

34.

Seim

R. W.

Spates

C. R.

(2009). The prevalence and comorbidity of specific phobias in college students and their interest in receiving treatment. Journal of College Student Psychotherapy, 24(1), 49–58. https://doi.org/10.1080/87568220903400302

35.

Siegel

Anderson

J. F.

Han

(2011). Very brief exposure II: The effects of unreportable stimuli on reducing phobic behavior. Consciousness and Cognition, 20(2), 181–190. https://doi.org/10.1016/j.concog.2010.09.003

36.

Siegel

Warren

Jacobson

Merritt

(2018). Masking exposure to phobic stimuli reduces fear without inducing electrodermal activity. Psychophysiology, 55(5). https://doi.org/10.1111/psyp.13045

37.

Siegel

Warren

Wang

Yang

Cohen

Anderson

J. F.

Murray

Peterson

B. S.

(2017). Less is more: Neural activity during very brief and clearly visible exposure to phobic stimuli. Human Brain Mapping, 38(5), 2466–2481. https://doi.org/10.1002/hbm.23533

38.

Siegel

Weinberger

(2009). Very brief exposure: The effects of unreportable stimuli on fearful behavior. Consciousness and Cognition, 18(4), 939–951. https://doi.org/10.1016/j.concog.2009.08.001

39.

Siegel

Weinberger

(2012). Less is more: The effects of very brief versus clearly visible exposure. Emotion, 12(2), 394–402. https://doi.org/10.1037/a0026806

40.

Stahl

Haaf

Corneille

(2016). Subliminal evaluative conditioning? Above- chance CS identification may be necessary and insufficient for attitude learning. Journal of Experimental Psychology, 145(9), 1107–1131. https://doi.org/10.1037/xge0000191

41.

Stinson

F. S.

Dawson

D. A.

Chou

S. P.

Smith

Goldstein

R. B.

Ruan

W. J.

Grant

B. F.

(2007). The epidemiology of DSM-IV specific phobia in the USA: Results from the national epidemiologic survey on alcohol and related conditions. Psychological Medicine, 37(7), 1047–1059. https://doi.org/10.1017/S0033291707000086

42.

Taschereau-Dumouchel

Cortese

Chiba

Knotts

J. D.

Kawato

Lau

(2018). Towards an unconscious neural reinforcement intervention for common fears. Proceedings of the National Academy of Sciences, 115(13), 3470–3475. https://doi.org/10.1073/pnas.1721572115

43.

Tottenham

Tanaka

J. W.

Leon

A. C.

McCarry

Nurse

Hare

T. A.

Marcus

D. J.

Westerlund

Casey

B. J. J.

Nelson

(2009). The NimStim set of facial expressions: Judgments from untrained research participants. Psychiatry Research, 168(3), 242–249. https://doi.org/10.1016/j.psychres.2008.05.006

44.

Tsuchiya

Koch

(2005). Continuous flash suppression reduces negative afterimages. Nature Neuroscience, 8(8), 1096–1101. https://doi.org/10.1038/nn1500

45.

Valuch

Mattler

(2019). Action priming is linked to visual perception in continuous flash suppression. Journal of Vision, 19(7). https://doi.org/10.1167/19.7.13

46.

van Dis

E. A.

Hagenaars

M. A.

Bockting

C. L.

Engelhard

I. M.

(2019). Reducing negative stimulus valence does not attenuate the return of fear: Two counterconditioning experiments. Behaviour Research and Therapy, 120(1). https://doi.org/10.1016/j.brat.2019.103416

47.

Wardenaar

K. J.

Lim

C. C.

Al-Hamzawi

A. O.

Alonso

Andrade

L. H.

Benjet

C. d.

Bunting

De Girolamo

Demyttenaere

Florescu

S. E.

Gureje

Hisateru

Huang

Karam

Kiejna

Lepine

J. P.

Navarro-Mateu

Oakley Browne

de Jonge

(2017). The cross- national epidemiology of specific phobia in the world mental health surveys. Psychological Medicine, 47(10), 1744–1760. https://doi.org/10.1017/S0033291717000174

48.

Yamashiro

Yamamoto

Mano

Umeda

Higuchi

Saiki

(2014). Activity in early visual areas predicts interindividual differences in binocular rivalry dynamics. Journal of Neurophysiology, 111(6), 1190–1202. https://doi.org/10.1152/jn.00509.2013

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.31 MB

0.07 MB