Date of Degree


Document Type


Degree Name



Cognitive Neuroscience


David Johnson

Subject Categories

Biological Psychology | Cognitive Neuroscience | Cognitive Psychology


Fear, Learning, Pavlov, Conditioning, Overtraining


Successful regulation of fear memories is a fundamental tenet to the exposure-based therapies often employed by mental health professionals for individuals with PTSD, phobias, and other anxiety disorders. Consequently, the efficacy of these treatment methodologies is largely dependent on the strength of the fear memory, as stronger memories are often characterized by an increased resistance to extinction and heightened fear recovery. However, there is little consensus within the scientific community regarding how to effectively maximize fear memory strength in human studies, and the literature exploring the impact of variability in acquisition parameters on memory strength is sparse. Here, we tested the effects of learning experience on memory strength by employing a within-subject Pavlovian fear conditioning paradigm over the course of three days with separate phases of acquisition, extinction, and fear recovery. Participants (N=44) were conditioned to two stimuli (colored squares), which were paired with an electric shock. While reinforcement rate was matched across both conditioned stimuli at 50%, learning experience was increased by 80% for one stimulus (CS+high) compared to the other (CS+low). Primary output measures were SCR, self-reported US expectancy and evaluation of the severity of the US; the latter two were combined to create a compound measure, subjective expected threat value (ETV). We found no difference in the SCRs between the two CS+s during acquisition, extinction, and fear recovery stages of the experiment, though we were able to identify a descriptive, but not statistical, difference in the US expectancy ratings, with participants consistently overestimating the reinforcement rate of the CS+high compared to the CS+low, suggesting that participants expectancy ratings are reflective of the number of CS-US pairings. The finding of no effect of learning experience on memory strength, as indexed by our implicit measure of SCR, calls into question the viability of this specific protocol as a reliable method for manipulating memory strength. One possible explanation for these null results is that learning reached its asymptotic limit for the CS+low and a ceiling effect prevented further learning from occurring for the CS+high (a true negative result). Alternatively, it is possible that learning experience did mediate differences in strength between the two memories, but the paradigm was not sufficiently sensitive to pick these differences up (a false negative result). For the explicit measure of US expectancy, participants showed some evidence of differentiating between the two CS+s, with a trend towards higher, and more inaccurate, estimates of the reinforcement rate for the CS+high compared to the CS+low, though these differences did not meet traditional statistical thresholds. These descriptive results are in contrast with standard learning theory and could merit further exploration.