Detecting Artificial Intelligence-Generated Personal Statements in Professional Physical Therapist Education Program Applications: A Lexical Analysis.

OBJECTIVE: The objective of this study was to compare the lexical sophistication of personal statements submitted by professional physical therapist education program applicants with those generated by OpenAI's Chat Generative Pretrained Transformer (ChatGPT). METHODS: Personal statements from 152 applicants and 20 generated by ChatGPT were collected, all in response to a standardized prompt. These statements were coded numerically, then analyzed with recurrence quantification analyses (RQAs). RQA indices including recurrence, determinism, max line, mean line, and entropy were compared with t-tests. A receiver operating characteristic curve analysis was used to examine discriminative validity of RQA indices to distinguish between ChatGPT and human-generated personal statements. RESULTS: ChatGPT-generated personal statements exhibited higher recurrence, determinism, mean line, and entropy values than did human-generated personal statements. The strongest discriminator was a 13.04% determinism rate, which differentiated ChatGPT from human-generated writing samples with 70% sensitivity and 91.4% specificity (positive likelihood ratio = 8.14). Personal statements with determinism rates exceeding 13% were 8 times more likely to have been ChatGPT than human generated. CONCLUSION: Although RQA can distinguish artificial intelligence (AI)-generated text from human-generated text, it is not absolute. Thus, AI introduces additional challenges to the authenticity and utility of personal statements. Admissions committees along with organizations providing guidelines in professional physical therapist education program admissions should reevaluate the role of personal statements in applications. IMPACT: As AI-driven chatbots like ChatGPT complicate the evaluation of personal statements, RQA emerges as a potential tool for admissions committees to detect AI-generated statements.

Recurrence Quantification Analysis of Personal Statements from Applicants to a Physical Therapy Doctoral Program: A Cross-Sectional Analysis.

PURPOSE: We examined lexical sophistication of written personal statements from physical therapy applicants to determine whether recurrence quantification analysis (RQA) indices distinguish higher- versus lower-scored candidates and correlate with other admissions variables. METHODS: Written personal statements were extracted from 152 applications, coded numerically, and analyzed with RQA. Along with other RQA indices, determinism-representing predictability of words and phrases-was quantified. A receiver operating characteristic (ROC) curve analysis was used to examine discriminative validity of RQA indices to distinguish between top-10 and bottom-10 candidates. Correlation coefficients between RQA indices and other admissions variables (grade point averages, standardized exam, behavioral interview, and cumulative admissions scores) were also examined. RESULTS: Determinism in personal statements was lower in top-scored (mean 7.38%) than bottom-scored candidates (mean 11.29%, p = 0.015), differentiated between them with 70% sensitivity (95% CI 34.8%-93.3%) and 100% specificity (95% CI 69.2%-100%), and correlated negatively with candidates' behavioral interview scores (r = -0.168, p = 0.039). DISCUSSION: The greater expressive lexical sophistication characterized by lower determinism in personal statements provides information about candidates' writing proficiency as a component of their communication abilities that may not otherwise be assessed in traditional admissions variables.

Estimation of manual wheelchair-based activities in the free-living environment using a neural network model with inertial body-worn sensors.

Shoulder pain is common in manual wheelchair (MWC) users. Overuse is thought to be a major cause, but little is known about exposure to activities of daily living (ADLs). The study goal was to develop a method to estimate three conditions in the field: (1) non-propulsion activity, (2) MWC propulsion, and (3) static time using an inertial measurement unit (IMU). Upper arm IMU data were collected as ten MWC users performed lab-based MWC-related ADLs. A neural network model was developed to classify data as non-propulsion activity, propulsion, or static, and validated for the lab-based data collection by video comparison. Six of the participants' free-living IMU data were collected and the lab-based model was applied to estimate daily non-propulsion activity, propulsion, and static time. The neural network model yielded lab-based validity measures ≥0.87 for differentiating non-propulsion activity, propulsion, and static time. A quasi-validation of one participant's field-based data yielded validity measures ≥0.66 for identifying propulsion. Participants' estimated mean daily non-propulsion activity, propulsion, and static time ranged from 158 to 409, 13 to 25, and 367 to 609 min, respectively. The preliminary results suggest the model may be able to accurately identify MWC users' field-based activities. The inclusion of field-based IMU data in the model could further improve field-based classification.

Application and Reliability of Accelerometer-Based Arm Use Intensities in the Free-Living Environment for Manual Wheelchair Users and Able-Bodied Individuals.

Arm use in manual wheelchair (MWC) users is characterized by a combination of overuse and a sedentary lifestyle. This study aimed to describe the percentage of daily time MWC users and able-bodied individuals spend in each arm use intensity level utilizing accelerometers. Arm use intensity levels of the upper arms were defined as stationary, low, mid, and high from the signal magnitude area (SMA) of the segment accelerations based on in-lab MWC activities performed by eight MWC users. Accelerometry data were collected in the free-living environments from forty MWC users and 40 sex- and age-matched able-bodied individuals. The SMA intensity levels were applied to the free-living data and the percentage of time spent in each level was calculated. The SMA intensity levels were defined as, stationary: ≤0.67 g, low: 0.671-3.27 g, mid: 3.27-5.87 g, and high: >5.871 g. The dominant arm of both MWC users and able-bodied individuals was stationary for most of the day and less than one percent of the day was spent in high intensity arm activities. Increased MWC user age correlated with increased stationary arm time (R = 0.368, p = 0.019). Five and eight days of data are needed from MWC users and able-bodied individuals, respectively, to achieve reliable representation of their daily arm use intensities.

Comparing supraspinatus to acromion proximity and kinematics of the shoulder and thorax between manual wheelchair propulsion styles: A pilot study.

BACKGROUND: Repetitive glenohumeral joint movement during manual wheelchair propulsion has been associated with shoulder pain in individuals with spinal cord injury. Clinical guidelines for shoulder health maintenance encourage semi-circular over arc propulsion to reduce loading frequency. This study aimed to determine the difference in estimated supraspinatus to acromion compression risk, and shoulder, thorax kinematics between (1) arc and semi-circular propulsion; and (2) self-selected and coached strategies. METHODS: Shoulder and thorax kinematics were captured during wheelchair propulsion, noting individually self-selected styles. Participants were then coached to perform the other style(s) of interest, arc and/or semi-circular. CT bone models of the humerus and scapula were animated using glenohumeral kinematics to estimate the minimum distance between the supraspinatus humeral attachment and the acromion. Compression risk was defined as the proportion of each propulsion phase where the minimum distance fell below 5 mm. Comparisons were made between conditions evaluating compression risk, minimum distances and kinematics at events throughout propulsion. FINDINGS: Ten individuals with spinal cord injury (9 male) participated. Arc and semi-circular propulsion did not significantly differ in compression risk or minimum distance across propulsion phases. Self-selected styles yielded lower compression risk and larger proximity values compared to coached styles. Glenohumeral horizontal abduction and thorax flexion differed between arc and semi-circular propulsion. Multiple glenohumeral and humerothoracic differences emerged between self-selected and coached conditions. INTERPRETATION: Supraspinatus compression was observed during both arc and semi-circular propulsion, suggesting risk may be unavoidable in this task. Self-selected styles yield less risk, likely related to coached style unfamiliarity.