Manipulating Emotions

Correlational Studies

As you most likely learned in other classes, correlational studies identify the significant direction between two variables. A positive correlation indicates both variables increase or both variables decrease. A negative correlation indicates one variable increases as the other variable decreases. A correlation close to zero suggests a relationship between two variables does not exist. As the correlation becomes closer to +1 or -1, the correlation becomes stronger. Remember, with correlational designs causation cannot be determined. Thus, when describing a correlation, the following terms are appropriate: correlated, related to, and associated with. When describing a correlation, the terms caused, led to, and influenced should be avoided. Finally, when explaining a correlational relationship, always indicate the direction – whether the two variables are positively or negatively correlated.

Experimental Studies

When we are looking to identify causation between two variables, we must conduct an experiment. To meet the requirements for an experiment, the study must randomly assign participants to one of the independent variable conditions. In emotions research, typically the independent variable conditions include eliciting one or more emotions, and sometimes including a control condition. Some possible independent variable conditions are in Table 2 below.

Table 2
Example Independent Variable Conditions

After eliciting the emotion in the independent variable conditions, research will typically measure changes in one or more of the emotion components.

When an independent variable is between-subjects, each participant experiences one of the independent variable conditions one time. For instance, we might measure skin conductance while 1/3 of our participants watch a sad clip, 1/3 watch a happy clip, and 1/3 watch a neutral clip. Each participant only completes one of the three conditions. Typically, in lab studies, emotions are elicited over several trials. Often, the first trial is a baseline trial that occurs before the manipulation. For within-subject independent variables, each participant experiences all the independent variable conditions. For instance, let’s say participants will be looking at 10 disgusting photos while we measure their heart rate. The independent variable conditions would be a baseline measure of heart rate, photo 1 heart rate, photo 2 heart, etc. Thus, this one independent variable includes 11 conditions completed by all participants!

How do we elicit emotions?

Emotions can be elicited with a variety of techniques. Some common methods to elicit emotions include video clips, music clips, scenarios, photos, scripted social interaction, and recall of emotion experience. A scripted social interaction occurred in Schachter and Singer’s (1962), where confederates were used to elicit negative or position emotions in participants. In recall studies, participants are asked to recall a time they felt a specific emotion and then rate component changes. In scenarios, participants will be provided a description of a fictional character’s emotional experience and then ask to rate measures related to this emotion experience. The emotion music elicits depends on pitch, tempo, and the chords played. For instance, minor chords typically elicit sad emotions, whereas major chords typically elicit positive emotions. Dissonant chords elicit highly arousing, negative emotions like anxiety and nervousness, whereas consonant chords elicit low arousal positive emotions like contentment and calmness. Low pitch sounds (like an animal growling) elicit negatively-valenced emotions, whereas high pitch sounds (like Mariah Carey hitting those high notes), elicits positively-valenced emotions. Finally, slow tempo elicits low arousal emotions and fast tempo elicits high arousal emotions. So, if a researcher wanted to elicit a high arousal positive emotion, she might play a song in a major chord with a fast tempo!