Effectiveness of Single-Case Experiments

Naman Magan Bhatnagar

Blog Post 1:

           This blog post aims to demonstrate how a growing pool of descriptive accounts of single-case experiments can be equally useful to psychology as research produced through the experimental method. To prove this, I aim to consider the effectiveness of Henry Molaison (HM’s) case of anterograde amnesia. I shall do so by (1) describing why the experimental method provides valuable research to psychology in the first place, (2) discussing the shortcomings of the experimental method, (3) addressing how descriptive & relational single-case experiments provide valuable research the experimental method fails to cover, and finally (4) covering the extent to which single-accounts can be considered useful. 

Experimentation is the only way one can ensure research is valid and reliable. This gives scientific claims the power to be trusted and thus be of value. To do so, as we’ve learnt in class, experimentation relies on random assignment and manipulation of the independent variable to thoroughly rule out confounding variables. Take the example of the development of Selective Serotonin Reuptake Inhibitors (SSRI’s). In order to test the efficacy of SSRI’s, “a total of 131 randomised placebo-controlled trials enrolling a total of 27,422 participants were included” (Jakobsen, 2017). For the effects of human trials to be considered valuable, scientists required a large sample size, a control & experimental group, etc. to consider this research valuable. Without such experimentation, confounds such as people-specific variations in drug effects may have skewed or altered the results emphasizing why one needs the experimental method to sift for valuable research.

But then, how effective would the accounts and inferences learnt from HM’s single-case experiments be of use? His case would fail the test of random assignment and would be utterly convoluted in confounds. Due to this, no matter how many epileptic cases similar to HM’s one might encounter, no ‘valuable’ inferences would be drawn as inferences may become more reliable while remaining to be internally invalid. However, a rising case of reliability may indicate a likely correlation, as pointed out by William Scoville: “Whenever the hippocampus and hippocampal gyrus were damaged bilaterally in these operations some memory deficit was found” (Scoville, 1957). This tested true with 8 other patients who demonstrated similar results in conditions similar to HM’s. Due to ethical reasons, such cases cannot be artificially set limiting the pool of potential candidates for research. Thus, given the fact that these 9 cases provided the same result in such a small pool of candidates, does the fact that these cases weren’t randomly assigned completely undermine the correlation presented? Let’s explore how such correlations can still prove to be valuable.

As more single-case experiments arise over time, more inferences and correlations about the workings of the brain gain power. Additionally, these cases help (1) open researchers to new opportunities and avenues of exploration, (2) add a piece to our understanding of the brain, and (3) help scientists and doctors better identify possible causes of ailments. Because of HM’s case, scientists were able to (1) better able to understand the “abnormal changes” (Salat, 2006) caused by surgery, particularly (2) supporting Milner and Scoville’s inferences about the role of the hippocampus in memory loss. In addition, as a result of a subsequent 10-year follow-up, (3) HM’s case brought up new evidence related to “hypertension” (Salat, 2006) causing significant changes in deep grey matter and white matter structures. Here, we see systematic testing, biologically backed inferences, mounting correlations, a growing case-study pool, and an overall better understanding of the brain all due to descriptive and relational research into single-case experiments which fail to constitute experimental research. Moreover, as the Professor said, every time we move towards scientifically backed inferences to try and classify disorders, we lose vital elements of each patient’s individual accounts. Thus, the more we tend towards trying to make casual statements using the experimental method, the less generalizable our work tends to get which ultimately reduces those very elements each patient provides. In contrast, descriptive accounts enable us to record, test, and note every possible inference to compare with other single-case experiments that come. In essence, they enable us to save more information than the experimental method which may spawn new fields of research that were previously undiscovered, as Salat did.

Yet, the major drawback of descriptive methods is that every perspective and inference is drawn out an experimenter’s own account which opens the doors to interpretational bias. This is where experimental research essentially triumphs. Experimental research, without such accounts, would be just as stunted as single-case experiments would be stunted without experimental research. Thus, both facets prove to be equally useful.

In conclusion, though descriptive accounts of single-case experiments inherit bias, they still prove to be mighty useful to expanding the field of psychology and better analysing patient-treatment, as HM’s case was to Clinical Psychology, Neuropsychology, and so many epileptic patients that shall need treatment after him.

References:

Jakobsen, J. C.-M. (2017). Selective serotonin reuptake inhibitors. BMC Psychiatry.

Salat, D. H. (2006). Neuroimaging H.M.: A 10-year follow-up examination. Hippocampus, 16(11), 936–945.

Scoville, B. M. ( 1957). LOSS OF RECENT MEMORY AFTER BILATERAL HIPPOCAMPUL LESIONS. J. Neurol. Neurosurg. Psychiat., 20, 11.











COMMENTS: (Prithvi Iyer)

The paper reads well. It is well structured and the aim of the write up is made clear at the start. There are a few sentence structuring issues that I thought might be good to point out so as to further tighten the writing. 

For instance:-

Due to this, no matter how many epileptic cases similar to HM’s one might encounter, no ‘valuable’ inferences would be drawn as inferences may become more reliable while remaining to be internally invalid.

This sentence could be made clearer by either breaking the sentence down, or not using the same word multiple times within the same sentence. Furthermore, the sentence begins with the claim that no “valuable inferences” can be drawn but the reasons for them are not clear as the sentence goes on to refer to validity and reliability while alienating the notion of “value” that began the sentence. This is just one instance. For the next post, just try to tighten the writing. But for the first post, this is a very good piece. Well done!