Browsing by Author "Nandakumar, Ratna"
Now showing 1 - 2 of 2
Results Per Page
Sort Options
Item Bayesian Prior Choice in IRT Estimation Using MCMC and Variational Bayes(Frontiers Media SA, 2016-09-27) Natesan,Prathiba; Nandakumar,Ratna; Minka,Tom; Rubright,Jonathan D.; Prathiba Natesan, Ratna Nandakumar, Tom Minka, Jonathan D.Rubright; Nandakumar, RatnaThis study investigated the impact of three prior distributions: matched, standard vague, and hierarchical in Bayesian estimation parameter recovery in two and one parameter models. Two Bayesian estimation methods were utilized: Markov chain Monte Carlo (MCMC) and the relatively new, Variational Bayesian (VB). Conditional (CML) and Marginal Maximum Likelihood (MML) estimates were used as baseline methods for comparison. Vague priors produced large errors or convergence issues and are not recommended. For both MCMC and VB, the hierarchical and matched priors showed the lowest root mean squared errors (RMSEs) for ability estimates: RMSEs of difficulty estimates were similar across estimation methods. For the standard errors (SEs), MCMC-hierarchical displayed the largest values across most conditions. SEs from the VB estimation were among the lowest in all but one case. Overall, VB-hierarchical, VB-matched, and MCMC-matched performed best. VB with hierarchical priors are recommended in terms of their accuracy, and cost and (subsequently) time effectiveness.Item Establishing severity levels for patient-reported measures of functional communication, participation, and perceived cognitive function for adults with acquired cognitive and language disorders(Quality of Life Research, 2022-12-27) Cohen, Matthew L.; Harnish, Stacy M.; Lanzi, Alyssa M.; Brello, Jennifer; Hula, William D.; Victorson, David; Nandakumar, Ratna; Kisala, Pamela A.; Tulsky, David S.Purpose: To empirically assign severity levels (e.g., mild, moderate) to four relatively new patient-reported outcome measures (PROMs) for adults with acquired cognitive/language disorders. They include the Communicative Participation Item Bank, the Aphasia Communication Outcome Measure, and Neuro-QoL’s item banks of Cognitive Function (v2.0) and Ability to Participate in Social Roles and Activities (v1.0). Method: We conducted 17 focus groups that comprised 22 adults with an acquired cognitive/language disorder from stroke, Parkinson’s disease, or traumatic brain injury; 30 care partners of an adult with an acquired cognitive/language disorder; and 42 speech-language pathologists who had experience assessing/treating individuals with those and other cognitive/language disorders. In a small, moderated focus-group format, participants completed “PROM-bookmarking” procedures: They discussed hypothetical vignettes based on PROM item responses about people with cognitive/language disorders and had to reach consensus regarding whether their symptoms/function should be categorized as within normal limits or mild, moderate, or severe challenges. Results: There was generally good agreement among the stakeholder groups about how to classify vignettes, particularly when they reflected very high or low functioning. People with aphasia described a larger range of functional communication challenges as “mild” compared to other stakeholder types. Based on a consensus across groups, we present severity levels for specific score ranges for each PROM. Conclusion: Standardized, stakeholder-informed severity levels that aid interpretation of PROM scores can help clinicians and researchers derive better clinical meaning from those scores, for example, by identifying important clinical windows of opportunity and assessing when symptoms have returned to a “normal” range.