China
Africa
Explore chapters and articles related to this topic
Solutions Using Machine Learning for Diabetes
Published in Punit Gupta, Dinesh Kumar Saini, Rohit Verma, Healthcare Solutions Using Machine Learning and Informatics, 2023
Jabar H. Yousif, Kashif Zia, Durgesh Srivastava
The purpose of the statistical evaluation of medical data is to express relations among its variables mathematically, for example, to analyze patients affected by diabetes (Y) over the years (X). Diabetes is the dependent variable to be defined, and the independent variable is over the years. There are three types of regression analyses: linear regression, logistic regression, and cox regression [25].
Survival Modeling I: Models for Exchangeable Observations
Published in Gary L. Rosner, Purushottam W. Laud, Wesley O. Johnson, Bayesian Thinking in Biostatistics, 2021
Gary L. Rosner, Purushottam W. Laud, Wesley O. Johnson
As we have encountered in several previous chapters, regression analyses are much more common in biomedical research than are exchangeable (or conditionally iid) sampling studies of a single homogeneous population. Time-to-event studies are no exception to this. The next chapter addresses regression in this context. Much of the current literature routinely includes regression of various types. We chose to separate the exchangeable data case in a separate chapter for pedagogical reasons. The new concepts that relate to censoring and to inference targets such as survival and hazard functions are easier to introduce without having to model how certain covariates might affect these inference targets, particularly in the presence of censoring. As most literature includes regression and treats the exchangeable data situation as a special case, we defer recommending readings until the end of the next chapter.
Basic stats
Published in O. Ajetunmobi, Making Sense of Critical Appraisal, 2021
As described above, a linear regression line is used to model the relationship between variables X and Y. This linear model tells us how Y changes with X, thereby making it possible to predict the value of Y for any given value of X. This linear model can be represented algebraically as a linear regression equation (Figure 1.15).
Working Alliance Inventory (WAI) and its relationship to patient-reported outcomes in painful musculoskeletal conditions
Published in Disability and Rehabilitation, 2023
Mary Beth Holmes, Amanda Scott, James Camarinos, Lee Marinko, Steven Z. George
There are several limitations to this study. Data were collected via a convenience sampling approach from a single outpatient clinic, where the patient population was relatively young, high functioning, and lower in pain compared to that seen in the general population. This may introduce some challenges in interpreting how outcomes were met or trends in change as the tools approach their relative ceiling. The small sample size limited multivariant statistical approaches. A limited number of variables were included in regression modeling to respect the limitations of the sample size. Several body region specific outcome measures were used in this study, converting outcome change scores to standardized z-scores was used to mitigate the impact of this potential limitation and allowed for statistical analysis using the entire sample. The convenience sample represented individuals with complete data sets which could further limit the generalizability of the findings, though we feel served the needs of this exploratory study. As noted previously, our study also did not attempt to compare intervention type or appropriateness of intervention type relative to the outcome or TA measured.
Spoken propositional idea density, a measure to help second language English speaking students: A multicentre cohort study
Published in Medical Teacher, 2022
Andrew M. Lunn, Daniel Matthias Bürkle, Rebecca Ward, Alice P. McCloskey, Adam Rathbone, Aaron Courtenay, Rachel Mullen, Andrea Manfrin
Logistic regression is a statistical technique used to determine the relationship between predictors, represented by independent variables, and predicted variables (dependent variables). The logistic regression aims to predict the odds (Exp (B)) of success (occurred) or failure (not occurred), representing the ratio of the probability (p) [odds = p/(1 − p)]. Its basic function is the logistic model (logit). Thus, the logarithm of the odds is represented by the logit function [logit p = In (p/1 − p)] for 0 < p < 1. In our study, we were dealing with dichotomous variables, therefore two binary logistic regression models were used. The independent dichotomous variables were the same for the two models, age, gender, and language. The dependent variables in the models were: LID score dichotomised and MID score dichotomised (<70% and ≥70% and above).
An Examination of Variables that Predict Turnover, Staff and Caregiver Satisfaction in Behavior-analytic Organizations
Published in Journal of Organizational Behavior Management, 2022
Daniel J. Cymbal, Sara Litvak, David A. Wilder, Gary N. Burns
Regression refers to a family of statistical methods used to describe the relationship between two or more variables. Specifically, regression procedures produce a line of best fit for a set of variables. Simultaneous multiple regression, used to analyze these data, is so named because all variables are entered into the regression equation at the same time (see Cohen, Cohen, West, & Aiken, 2013 for a thorough explanation). Additionally, when describing the relationship between variables, it may be prudent to investigate the degree to which one variable predicts changes in the criterion variable relative to others. For these data, we use dominance analysis, a method which estimates a predictor variable’s relative contribution by determining its average contribution across all possible combinations of the regression model (see Tonidandel and LeBreton (2011) for a discussion on appropriate methods to determine a variable’s relative importance).