Date Thesis Awarded
Honors Thesis -- Open Access
Bachelors of Arts (BA)
This research investigates how to measure affected speech of patients with schizophrenia by analyzing how they orally describe a picture compared to controls. Currently, there is no single clear set of criteria for recognizing disorganized speech. By working with a standard set of parameters (descriptions of a single picture) we can find patterns of speech that differ between the two groups.
68 patients and 78 controls were asked to describe a line drawing of a beach scene as completely as possible for a period of 2 minutes. These picture descriptions were analyzed on the basis of i) ɴᴀʀʀᴀᴛɪᴠᴇ sᴛʀᴜᴄᴛᴜʀᴇ, which examines how patients and controls transition between topics; ii) ʀᴇʟᴀᴛɪᴏɴsʜɪᴘs, which investigates how subjects construct connections between objects in the picture; and iii) ᴄᴏᴍᴘʟᴇᴛᴇɴᴇss, which investigates whether patients and controls differ in how detailed their respective descriptions are. The major findings of these analyses showed that patients use fewer complex transitions than controls when switching topics (that is, transitions that provide additional information, rather than basic ones like “and”), make fewer connections between objects in the picture than controls, and give less complete picture descriptions with fewer details.
These results are further expanded upon in a series of follow-up analyses, showing alogia’s effects on these findings and correlations between these linguistic measurements and psychiatric interview tools clinicians use to rate symptom severity in schizophrenia. In particular, the word counts of participants’ picture descriptions had a significant impact on the above findings, implying that alogia rather than disorganized speech may account for these differences. Additionally, correlations were found between the linguistic measurements of this paper and clinical measurements via psychiatric interviews for disorganized speech in patients participating in the study, but not with their ratings of alogia. Finally, this paper also makes recommendations about prompting for additional speech in picture description tasks, based on evidence showing that discourse patterns do not change significantly when participants are encouraged to speak more.
Metzger, Celia, "Linguistic Measures of Symptomatology in Schizophrenia" (2021). Undergraduate Honors Theses. William & Mary. Paper 1605.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.