Inhibition of vicariously learned fear in children using positive modeling and prior exposure.

Experiment 1

One reason why some children do not develop fear responses during a fear-related vicarious learning event may be because they do not interpret the US (the model’s response to the stimulus) as intense enough for learning to occur; for example, if they do not believe that the scared model is a reliable source of information about how threatening the stimulus is. One way to test this would be to use information to manipulate children’s interpretation of models’ responses. Psychoeducation may be one method to reduce negative learning expectancies. It can be used to help children and their families understand how anxiety develops, how it is maintained, and how it is treated, as well as to teach children to recognize emotional and physiological signs of anxiety and develop coping strategies. Psychoeducation is one of the tools used by clinicians to prevent ( Barrett et al., 2014 ) or treat childhood disorders by, for example, correcting threat-related myths and false assumptions about a feared stimulus, as well as catastrophic expectancies ( Davis, Ollendick, & Öst, 2009 ). For example, it can successfully reduce spider fear in 8- to 10-year-olds ( Leutgeb, Schaider, & Schienle, 2012 ). If psychoeducation could be used to reduce how threatening children interpret a model’s fearful responses to be (i.e., reduce the intensity of the US), associative learning models would predict a reduction in the learnt fear response to CSs that become associated with the US. This devaluation of the US could explain the resilience that some children show to fear-related vicarious learning and clarify the mechanisms underpinning psychoeducation as a prevention and treatment method. Thus in Experiment 1, children were divided into three groups and one group was given information about how sometimes individuals may respond to a stimulus with fear even though the stimulus is not actually threatening or dangerous. A specific example was given in which a child became afraid of a nondangerous stimulus after observing an adult acting afraid of it. Between the ages of 4 to 6 years children typically begin to develop an understanding that a person’s emotional responses to a situation are determined, at least in part, by their beliefs ( Pons, Harris, & de Rosnay, 2004 ). Consequently, it was predicted that this information would lead children (aged 7 to 9 years) to devalue models’ fear responses (US) to stimuli (CSs) during vicarious learning, making the outcome less threatening and hence reduce fear learning (CR). A second group of children were given nonthreatening factual information about the CS to create neutral expectancies. A third (control) group was given irrelevant information, presented in a similar manner and format to the other two conditions, but that was not related to the CS or US. Given predictions from conditioning models, and that vicarious learning has been argued to share the same characteristics ( Field & Purkis, 2011 ) it was expected that psychoeducation and nonthreatening information would inhibit vicarious fear learning compared to the control group.

Fifty-six (30 boys, 26 girls) children aged between 7.83 and 9.83 years ( M = 8.63 years, SD = 0.53) were recruited from a primary school in West Sussex, U.K. Parents were not asked to provide sociodemographic information but school records showed that children attending this village school were predominantly of White British background and fluent English speakers. The number of socioeconomically disadvantaged children attending the school was well below the national average. The school and parents/caregivers gave informed opt-in consent for children to take part and children gave their verbal assent. Children were randomly allocated to three information groups. There were initially 61 children but one child did not complete the entire procedure and four others were excluded because they stated at the end of the experiment that they were familiar with the animals. Thus there were 20 children in the psychoeducation group, 17 in the factual information group, and 19 children in the control group. A one-way independent ANOVA (Brown-Forsythe) found no significant difference between trait anxiety scores in the three groups (psychoeducation: M = 36.45, SD = 8.24; factual information: M = 34.71, SD = 4.04; control: M = 33.53, SD = 3.85), F (2, 36.97) = 1.30, p = .28. There was also no significant difference in mean ages in the three groups (psychoeducation: M = 8.58 years, SD = 0.56; factual information: M = 8.65 years, SD = 0.49; control: M = 8.66 years, SD = 0.55), F (2, 53) = 0.11, p = .90.

Faces (USs)

Emotional face stimuli consisted of 20 portrait photographs (400 × 513 pixels) taken from the NimStim Set of Facial Expressions ( Tottenham et al., 2009 ). There were five female and five male faces displaying happy facial expressions, and five female and five male faces displaying fearful facial expressions.

Fear Beliefs Questionnaire (FBQ)

A computerized version of the 21-item version of the FBQ ( Field & Lawson, 2003 ) was used and consisted of seven identical questions (four reverse-scored) for each of the three animals: quoll, quokka, and cuscus. Items are designed to measure children’s fear-related beliefs for the animals; for example, “Would you keep your distance if you saw a quoll/quokka/cuscus?” or “Would you be happy to have a quoll/quokka/cuscus for a pet?” (reverse-scored). Children responded on a 5-point Likert-type response scale (0 = no, not at all ; 1 = no, not really ; 2 = do not know/neither ; 3 = yes, probably ; 4 = yes, definitely ). Scores were averaged across items creating a mean fear belief score for each animal ranging from 0–4, with higher scores indicating higher levels of fear beliefs. Internal consistencies (Cronbach’s alpha) were high before learning: α = .74 (Cuscus subscale), .79 (Quokka subscale) and .82 (Quoll subscale); and after learning, αs = .89, .90 and .83, respectively.

Nature Reserve Task (NRT)

An adaptation of Field and Storksen-Coulson’s (2007) NRT was used to measure children’s avoidance preferences for the animals. The NRT consisted of a circular wooden board (diameter = 37.5 cm) covered in green material that children were asked to imagine was a nature reserve where the three animals lived. A rectangular 2-D cardboard photograph of each animal (3.4 × 3.4 cm with a 1-cm base as a stand) was placed, one at a time, at a specific point on the circumference of the board. The order of presentation of each animal was counterbalanced across children. Children were asked to imagine they were visiting the nature reserve and place a male (for boys) or a female (for girls) Lego figure (1.5 × 3.5 × 0.8 cm) where they would most like to be. The distance (in mm) between the center of the figure and the center of the animal picture was measured for each of the three animals to indicate children’s approach or avoidance preferences for that animal.

Factual information

A similarly formatted nine-page leaflet (A5 size) with nine color pictures of quolls, cuscuses, and quokkas was created to communicate factual information about the marsupials. For the purpose of the information leaflet, fairly neutral and nonthreatening commonalities between the three animals such as size, nocturnal lifestyle, diet, and predators were emphasized. For example, “Cuscuses, Quolls and Quokkas are nocturnal, which means that they sleep during the day and search for food at night. They have very good eyesight to help them see in the dark” and “Cuscuses, Quolls and Quokkas are quite small—about half a meter long. They like to eat fruit, nuts, leaves, insects and smaller animals, but not people!”

Unrelated information (control)

A third nine-page leaflet (A5 size) with nine color pictures was created to communicate information about ancient Greek gods and goddesses. This was a topic entirely unrelated to the CS, US, or experimental hypothesis and so served as a control manipulation. The style was comparable to the other leaflets with color cartoon pictures and informative text; for example, “Ancient Greeks believed that gods and goddesses watched over them. The gods were like humans, but immortal (they lived forever) and much more powerful” and “Zeus was the king of the Greek gods, and ruled all of the heavens. He threw thunderbolts to punish anyone who disobeyed him.”

Vicarious learning

Vicarious learning consisted of a total of 30 animal-face pairing trials presented in random order. Each child saw 10 trials in which one of the animal pictures was presented together with scared face pictures (fear-paired trials), 10 trials in which one of the animals was seen with happy faces (happy-paired trials), and 10 trials of the third animal alone on the screen (unpaired control trials). The type of emotional face (fear, happy, or none) children saw with each particular animal (quoll, quokka, or cuscus) was determined by the counterbalancing group they were in. A single animal–face paring trial was 2-s long and began with a randomly chosen animal picture being presented alone on the screen for 1 s on a randomly determined side of the screen. The animal continued on the screen for a further 1 s while the emotional face was presented on the opposite of the screen. Unpaired animals trials were identical except that no face appeared and the animal remained on the screen alone for 2 s. Between each trial there was a randomly determined 2-s to 4-s intertrial interval (see, e.g., Askew & Field, 2007 ; Askew & Field, 2008 ).

Procedure

Children first completed the STAIC and the FBQ. The FBQ and later vicarious learning procedure were automated on a Dell Inspiron 1300 laptop computer with a 13-inch monitor running Windows XP, using software custom-written by the fourth author in VisualBasic.net with ExacTicks 1.10 ( Ryle Design, 1997 ). Following the FBQ, children received one of three types of information depending on the group they were randomly allocated to: a) psychoeducation; b) factual information; or c) unrelated information (control). To ensure that participants continued to engage fully with the material, children were asked whether they would prefer to read the material aloud to the experimenter, or for the experimenter to read it to them. Next, children in each information group were randomly assigned to one of three counterbalancing groups and watched the automated vicarious learning procedure on the computer screen. Following this, children completed the second FBQ and then the NRT.

All children were fully debriefed at the end of the study and the aims were explained in age-appropriate language. Children were given the opportunity to ask any questions about the experiment and all questions were answered by the researcher. Factual information about the animals was given to the children both verbally and in a printed leaflet.