The authors have declared that no competing interests exist.
‡ DD and RL are joint senior authors on this work.
This study examines the impact of a visual representation of a secure base (i.e. a secure base prime) on attenuating experimentally produced anger and anxiety. Specifically, we examined the assuaging of negative emotions through exposure to an image of a mother-infant embrace or a heterosexual couple embracing. Subjects seated at a computer terminal rated their affect (Pre Affect) using the Affect Adjective Checklist (AAC) then listened to two sets of intense two person conflicts. After the first conflict exposure they rated affect again (Post 1 AAC). Following the second exposure they saw a blank screen (control condition), pictures of everyday objects (distraction condition) or a photo of two people embracing (Secure Base Prime condition). They then reported emotions using the Post 2 AAC. Compared to either control or distraction subjects, Secure Base Prime (SBP) subjects reported significantly less anger and anxiety. These results were then replicated using an internet sample with control, SBP and two new controls: Smiling Man (to control for expression of positive affect) and Cold Mother (an unsmiling mother with infant). The SBP amelioration of anger and anxiety was replicated with the internet sample. No control groups produced this effect, which was generated only by a combination of positive affect in a physically embracing dyad. The results are discussed in terms of attachment theory and research on spreading activation.
Several studies have established connections between attachment insecurity and anger [
Individuals whose history of experiences with caretakers has led them to lack confidence in their attachment figures are most prone to perceiving threats in close relationships and to responding with dysfunctional anger [
Bowlby proposed that individuals could shift in their attachment security over the lifespan as a function of corrective interpersonal experiences, such as relationships with significant others who are supportive in times of stress. Moreover, he saw a primary goal of therapy to be the restoration of deficits caused by early insecure attachment [
Shaver and Mikulincer [
There is other evidence that priming attachment security can have an ameliorative effect on selected outcome variables. For example, Pierce and Lydon [
The main explanation for semantic, as well as affective, priming has been spreading activation, a mechanism based on network models of memory [
Attachment security predicts affective responses to written anger-related scenarios [
Ethics approval for human participants was provided by the University of British Columbia Research Ethics Board (IRB) in the Department of Psychology. The UBC Department of Psychology IRB reviewed and approved this study. All participants provided written and/or electronic consent. All participants were required to read and approve the electronic/written consent statements before proceeding. No participants were allowed to participate in this study without consent. Automated pre-screening measures were also in place to verify that the participants' ages were above 18 years. No children were enrolled in this study.
Data was collected from 686 voluntarily participating university students of which 26.4% and 73.6% were male and female respectively. The students ranged in age from 18 to 59 with a mean age of 20.4 (
The internet sample encompassed participants from the United States. Of the 278 subjects that partook in the online studies, 58.6% were female and 41.4% were male. They ranged in age from 18 to 69 with a mean age of 34.6 (
The audios were credibly depicted by professional actors in an unscripted manner and recorded digitally. Audio conflicts involving outside scenarios, such as parking lots etc., had suitable background sounds inserted for enhanced authenticity. The actors were instructed to initiate the argument at a low to moderate level and then after a brief period, escalate the dispute by raising their voices, become verbally abusive and emotionally intense. All scenarios were of similar length and quality with respect to the emotional style, verbal content and emotional intensity. The conflict scenarios were selected from pre-testing for effectiveness (ability to generate emotional reactions as measured by the AAC—see above). Twenty pairs of counterbalanced scenarios were used in the study—these involved a variety of the issues described above as well as various combinations of gender of the persons depicted. Pairs of scenarios (conflict1, conflict 2) were used in a counter-balanced design because we initially intended to contrast affective reactions to stranger vs intimate conflict. However, no significant differences in were found between intimacy, stranger and family scenarios in their ability to generate affect (Anger, Anxiety and general AAC scores), hence, for purposes of data analysis all types of conflict are treated as commensurate. In effect, respondents received two exposures to conflict scenarios before differential exposure to SBP or control/distraction images.
Images were presented randomly to Ss in each of the experimental conditions. Each S was exposed to two conflicts and then saw only one image. In order to ensure that anger and anxiety were sufficiently elevated over pre-test levels to permit measurable reductions through SBP, we exposed Ss to two conflict scenarios. Control subjects saw a blank screen for 30s after each conflict and were asked to reflect on the conflict. Secure base prime (SBP) images consisted of either a mother with her infant or a heterosexual couple embracing. The distraction images were pictures of either a stapler, table, whistle or a musical note. These were used to test an alternative explanation for Secure Base Prime effects; namely that they served to distract Ss from aversive emotions. The internet sample had additional controls: a picture of a smiling Caucasian male (SM—Smiling Man) to test whether expression of positive emotion produced the same effect as SBP or a picture of a non-smiling mother holding an infant (CM—Cold Mother) to test whether mother-infant photos without positive affect could produce SBP.
Participants were instructed to sit in front of computer terminals presented with a full screen view displaying one of the three computer based experiments: a control group without any pictures, the Secure Base Prime picture group or the distraction picture group. As seen in
Control, SBP (Secure Base Prime), and distraction experiments. Experimental differences between the groups occur after Conflict Audio 2 as indicated by the red (SBP picture) and yellow boxes (distraction picture). AAC = Affective Adjective Checklist (Modified: 16 item).
The internet study used the same system that was used with the university sample albeit with some additional safeguards to help ensure uniformity in experience. The university study was conducted under controlled laboratory conditions to ensure that the experience was generally uniform for all subjects. Uniformity was maintained in the internet sample using participant screening, timing controllers on the study itself and pre-participation functionality checks. Anonymity was maintained by using two separate and unlinked databases. The participation experience was visually identical to that of the university group with the exception of additional visual instructions given in lieu of study invigilators reading certain instructions.
The percentage of mean differences of (Post1 AAC—Pre AAC) and (Post2 AAC—Post1 AAC) were calculated using this formula:
To check that Ss were attending to the stimuli and assessment questions, subjects were presented with five randomly inserted screening questions throughout the study. The screening questions depicted mostly impossible situations (e.g. “I have swum across a major ocean in the last year” or “a coin was recently minted with my image by the Federal Mint”). Subjects answering these questions could choose from “Strongly Agree (1)” to “Strongly Disagree (4)” on a four-point scale. Those scoring less than 4 on any of the five screening questions were invalidated. Initially, 811 participants’ data were recorded. 15.4% (n = 125) were deemed invalid as per our screening criteria which left 686 valid subjects that were analyzed for this study.
ANOVA was used to determine if there were differences between the experimental groups.
As with the pre-test, no significant differences were found between conflict scenarios tapping intimacy, family or stranger themes in either the Post1 AAC minus Pre AAC (i.e. Post1 AAC–Pre AAC) or Post2 AAC minus Post1 AAC (i.e. Post2 AAC—Post1 AAC) scores. Since these scenarios had been selected for maximum effectiveness (in terms of generating emotion), this was not surprising.
Post exposure self-report anger and anxiety were calculated by assessing each Post measure's increase in rating above the Pre measure. For the university sample,
ACC Sub Score | Experimental Group | Mean Difference % | Sig† | |
---|---|---|---|---|
Control (n = 222) | 8.09 | 15.01 | 0.33 | |
SBP (n = 284) | 7.98 | 15.65 | ||
Distraction (n = 180) | 10.02 | 15.77 | ||
Control (n = 222) | 2.68 | 14.57 | 0.01 | |
SBP (n = 284) | -0.51 | 15.14 | ||
Distraction (n = 180) | 3.06 | 13.14 | ||
Control (n = 222) | 2.36 | 19.02 | 0.51 | |
SBP (n = 284) | 2.96 | 20.19 | ||
Distraction (n = 180) | 4.54 | 17.50 | ||
Control (n = 222) | 0.06 | 14.64 | <0.001 | |
SBP (n = 284) | -5.62 | 15.59 | ||
Distraction (n = 180) | -1.09 | 15.26 |
Table Notes
† ANOVA
SBP: Secure Base Prime, SM:”Smiling Man”, CM:”Cold Mother and Child”.
SBP: Secure Base Prime, SM:”Smiling Man”, CM:”Cold Mother and Child”.
Most subjects reported low to moderate affect after hearing the conflict audios. To ascertain whether the SBP effect would occur for stronger levels of emotion, we chose the top 20 percent of subjects in each of the control and Secure Base prime conditions based on their Pre scores on Anger and Anxiety. The highest scorers in the top 20 percent were calculated separately for Anger and Anxiety scores. There were no statistical differences in the Post1 ACC—Pre ACC scores between the SBP and control group in either Anger or Anxiety whereas significant statistical differences were found between the SBP and control group in Post2 ACC—Post1 ACC scores (
ACC Sub Score | Experimental Group | Mean Difference % | Sig | |
---|---|---|---|---|
Control (n = 60) | 6.88 | 18.68 | 0.54 | |
SBP (n = 76) | 9.87 | 23.13 | ||
Control (n = 60) | 2.85 | 16.02 | 0.039 | |
SBP (n = 76) | - 4.00 | 20.40 | ||
Control (n = 55) | -7.65 | 18.94 | 0.91 | |
SBP (n = 75) | -7.11 | 24.17 | ||
Control (n = 55) | 1.21 | 13.77 | 0.001 | |
SBP (n = 75) | - 8.89 | 17.16 |
We again hypothesized a difference due to SBP in the Post2 ACC—Post1 ACC scores. There were no statistical differences at Post1 ACC—Pre ACC among any of the groups in either Anger or Anxiety (
ACC Sub Score | Experimental Group | Mean Difference % | Sig |
|
---|---|---|---|---|
Control (n = 72) | 4.51 | 15.87 | 0.12 | |
SBP (n = 69) | 8.99 | 17.19 | ||
Smiling Man (SM) (n = 67) | 11.26 | 23.14 | ||
Cold Mother (CM) (n = 70) | 5.89 | 15.87 | ||
Control (n = 72) | 3.82 | 14.51 | <0.001 | |
SBP (n = 69) | -4.53 | 16.64 | ||
Smiling man (SM) (n = 67) | 3.05 | 17.29 | ||
Cold Mother (CM) (n = 70) | 6.31 | 14.18 | ||
Control (n = 72) | 3.01 | 22.53 | 0.79 | |
SBP (n = 69) | 6.16 | 22.07 | ||
Smiling Man (n = 67) | 6.09 | 28.12 | ||
Cold Mother (CM) (n = 70) | 6.19 | 16.01 | ||
Control (n = 72) | 3.65 | 19.27 | <0.001 | |
SBP (n = 69) | -7.55 | 17.50 | ||
Smiling Man (SM) (n = 67) | -1.49 | 19.10 | ||
Cold Mother (CM) (n = 70) | 2.98 | 11.44 |
Table Notes
† = ANOVA
Two control groups were added to the repertoire in the Internet Group. The first additional distraction control was an image of a partially “Smiling Man” (SM) and the second was that of an affectively flat woman with a young child or “Cold Mother” (CM). These images were placed in the same experimental position as the SBP image. We again hypothesized that there would be no differences between the groups in the Anger and Anxiety Post1 ACC—Pre ACC differences and this was indeed the case (
A general examination of
In this study, we increased self-reported anger and anxiety by exposing subjects to audio recordings of angry interpersonal conflicts. We then tested whether subsequent exposure to a secure base prime (a picture of two people embracing) would ameliorate both anger and anxiety. Introduction of a secure base prime did significantly reduce anger and anxiety responses of subjects exposed to the prime compared to responses of subjects in the control (no image) and or other control conditions (Smiling Man, Cold Mother) as indicated by significant between group differences in Post 2 AAC—Post 1 AAC Anger and Anxiety scores. This effect was not simply produced in minimal anger/anxiety conditions; it was replicated with High Scorer groups. It was also replicated with an internet sample. It was not due to distraction. There were no significant differences in Post 2 AAC—Post 1 AAC Anger and Anxiety scores between distraction and control condition subjects. It was produced by an image of a mother and infant embracing or by a heterosexual couple embracing. In each image, positive emotion is displayed. Control pictures used for distraction and/or positive emotion did not produce a SBP effect. An image showing positive emotion
A number of emotion inducing still photos have been used in emotion research [
We were interested in self-reported ratings of specific emotions rather than the two dimensions of pleasure and arousal studied by the IPPS studies, hence, we used a self-report scale similar to that used by Rottenberg et al. [
Mikulincer and Shaver [
The basis of the SBP effect in attachment theory is as follows: The attachment behavioural system responds to sources of threat with actions designed to generate proximity to a secure base [
Klauer and Musch [
This study used subliminal primes of threat related words or neutral words. The threat primes lead to faster identification of subsequent proximity related words than did the neutral prime. The effect was realized only on proximity related words. As Shaver and Mikulincer [
In this study we replicated the SBP effect, found initially in a university sample, with an internet sample. Furthermore, we did not merely use the internet sample to assess opinions or attitudes but as an experimental sample with random allocation of subjects to conditions, all run over the internet. Subjects were presented with the same stimuli and assessed with the same measures in both the university and Internet samples. The effects in terms of emotion measures and the SBP effect were identical. In our view, online experiments present a viable alternative to conventional reliance on university student samples and has an advantage in providing a more demographically heterogeneous sample, thus increasing ecological validity.
One could raise a question as to whether demand characteristics were present in the current studies—i.e. were the Ss cued to respond by the SBP picture or the control pictures? However, the studies were a between-subject design and the participants had only information about the two conflicts they had heard and the subsequent picture post conflict 2. They were asked a battery of “perceptions of conflict” questions (e.g. “Who started the conflict? Who escalated the conflict”?) in addition to the Affect Adjective Checklist. These would have further masked the intent of the study. They had to complete the Affect Adjective Checklist (AAC) with its' 16 descriptors of emotion. It's hard to see how, on the basis of this limited information, they could have adduced which descriptors were the focus of our study (e.g. Why not humiliated, helpless or sad as opposed to anger or anxiety?) but they responded differently on these other descriptors. Furthermore, even on the anger and anxiety measures, the results differed as shown above, more consistent with a veridical response and a subtle difference than a rote response to demand characteristics.
In the current studies we exposed Ss to secure base primes solely in the Post 2 position–after witnessing two conflicts. Future work should examine participants in the Post 1 position and attempt to assess prophylactic effects from SBP. Also, in the current study we examined the SBP effect as a main effect independent of Ss dispositional "attachment style". Future work should examine the interaction of the SBP with attachment style [
In the present study, exposure to intense two-person conflict, led to increases in self-reported emotion of anger and anxiety in both subject samples. Bowlby described the attachment behavioural system as activated by any threat and de-activated by contact with a secure base. From this theoretical perspective, our present conflicts may have generated symbolic threats or else conflict is experienced as threatening and the Secure Base Prime images activated memories of attachment security and hence, reduction of negative emotion. We used an image in this study to ensure a standardized SBP, however, future work could examine whether exposure to a SBP image at the beginning of the study, with instructions to imagine the image later (i.e. after conflict) could produce a SBP effect, acting in effect, as a positive introject for the participant. The present study suggests that therapeutic benefits may accrue from guided visualization of the type of images used in this study; that they could be combined with relaxation and breathing techniques to maximize stress reduction, generate positive affect and reduce emotional turbulence [