Detection of Insincere Grips: Multivariate Analysis Approach Dr Bhoomiah Dasari University of Southampton Southampton SO17 1BJ United Kingdom

Slides:



Advertisements
Similar presentations
High Resolution studies
Advertisements

Statistics for Quantitative Analysis
Quantitative Methods Topic 5 Probability Distributions
Aron, Aron, & Coups, Statistics for the Behavioral and Social Sciences: A Brief Course (3e), © 2005 Prentice Hall Chapter 6 Hypothesis Tests with Means.
Decision Analysis and Its Applications to Systems Engineering The Hampton Roads Area International Council on Systems Engineering (HRA INCOSE) chapter.
Client Assessment and Other New Uses of Reliability Will G Hopkins Physiology and Physical Education University of Otago, Dunedin NZ Reliability: the Essentials.
Validity and Reliability
Determining and Forecasting Load Reductions from Demand Side Programs September 11-12, 2007 Presented by: Bill Bland, V.P. Consulting, GoodCents Liza Thompson,
Describing Data: Measures of Dispersion
1 Eloise E. Kaizar The Ohio State University Combining Information From Randomized and Observational Data: A Simulation Study June 5, 2008 Joel Greenhouse.
1  1 =.
Confidence Intervals Chapter 9.
C82MST Statistical Methods 2 - Lecture 2 1 Overview of Lecture Variability and Averages The Normal Distribution Comparing Population Variances Experimental.
Lecture 2 ANALYSIS OF VARIANCE: AN INTRODUCTION
Developing Study Skills and Research Methods Unit Leader: Dr James Betts Unit Code: HL20107
Quantitative Methods Lecture 3
Inference in the Simple Regression Model
7.1 confidence Intervals for the Mean When SD is Known
Biostatistics Unit 5 Samples Needs to be completed. 12/24/13.
Psychometrics 101: Foundational Knowledge for Testing Professionals Steve Saladin, Ph.D. University of Idaho.
Chapter 4: Basic Estimation Techniques
Hypothesis Testing Goal: Make statement(s) regarding unknown population parameter values based on sample data Elements of a hypothesis test: Null hypothesis.
Lecture 14 chi-square test, P-value Measurement error (review from lecture 13) Null hypothesis; alternative hypothesis Evidence against null hypothesis.
Measurement, Evaluation, Assessment and Statistics
Neil H. Schwartz, Ph.D. Psych 560
How would you explain the smoking paradox. Smokers fair better after an infarction in hospital than non-smokers. This apparently disagrees with the view.
Matching in Case-Control Designs EPID 712 Lecture 13 02/23/00 Megan O’Brien.
Chapter 7: The Distribution of Sample Means
COURSE: JUST 3900 TIPS FOR APLIA Chapter 7:
Chapter 4 Inference About Process Quality
Comparison of 2 Population Means Goal: To compare 2 populations/treatments wrt a numeric outcome Sampling Design: Independent Samples (Parallel Groups)
Statistical Sampling.
Statistical vs. Practical Significance
Understanding p-values Annie Herbert Medical Statistician Research and Development Support Unit
Lecture 3 Validity of screening and diagnostic tests
Healey Chapter 7 Estimation Procedures
Statistical Analysis SC504/HS927 Spring Term 2008
Estimating a Population Mean When σ is Known: The One – Sample z Interval For a Population Mean Target Goal: I can reduce the margin of error. I can construct.
Confidence Intervals with Proportions
Summary Statistics When analysing practical sets of data, it is useful to be able to define a small number of values that summarise the main features present.
Correlation and Linear Regression
1 Introduction to Inference Confidence Intervals William P. Wattles, Ph.D. Psychology 302.
EPIDEMIOLOGY AND BIOSTATISTICS DEPT Esimating Population Value with Hypothesis Testing.
Model and Variable Selections for Personalized Medicine Lu Tian (Northwestern University) Hajime Uno (Kitasato University) Tianxi Cai, Els Goetghebeur,
Independent Sample T-test Often used with experimental designs N subjects are randomly assigned to two groups (Control * Treatment). After treatment, the.
Today Concepts underlying inferential statistics
Chapter 9 Flashcards. measurement method that uses uniform procedures to collect, score, interpret, and report numerical results; usually has norms and.
Inferential Statistics
Chapter 12 Inferential Statistics Gay, Mills, and Airasian
Determining Sample Size
Evidence Based Medicine
Ergonomics Maximum Voluntary Effort By: Group 10 - Marcus, Allan, Matt, Andre, and Lance.
Statistics & Biology Shelly’s Super Happy Fun Times February 7, 2012 Will Herrick.
Education Research 250:205 Writing Chapter 3. Objectives Subjects Instrumentation Procedures Experimental Design Statistical Analysis  Displaying data.
Comparing two sample means Dr David Field. Comparing two samples Researchers often begin with a hypothesis that two sample means will be different from.
University of Ottawa - Bio 4118 – Applied Biostatistics © Antoine Morin and Scott Findlay 08/10/ :23 PM 1 Some basic statistical concepts, statistics.
Correlational Research Chapter Fifteen Bring Schraw et al.
EVIDENCE ABOUT DIAGNOSTIC TESTS Min H. Huang, PT, PhD, NCS.
Evaluating Results of Learning Blaž Zupan
Three Broad Purposes of Quantitative Research 1. Description 2. Theory Testing 3. Theory Generation.
Stats Lunch: Day 3 The Basis of Hypothesis Testing w/ Parametric Statistics.
Correlation & Regression Analysis
Sampling Fundamentals 2 Sampling Process Identify Target Population Select Sampling Procedure Determine Sampling Frame Determine Sample Size.
1 Probability and Statistics Confidence Intervals.
The inference and accuracy We learned how to estimate the probability that the percentage of some subjects in the sample would be in a given interval by.
Educational Research Inferential Statistics Chapter th Chapter 12- 8th Gay and Airasian.
NURS 306, Nursing Research Lisa Broughton, MSN, RN, CCRN RESEARCH STATISTICS.
Simulation-based inference beyond the introductory course Beth Chance Department of Statistics Cal Poly – San Luis Obispo
Evaluating Results of Learning
Product moment correlation
Presentation transcript:

Detection of Insincere Grips: Multivariate Analysis Approach Dr Bhoomiah Dasari University of Southampton Southampton SO17 1BJ United Kingdom

Introduction A standardized grip strength assessment can provide us with quantifiable and objective information on clients hand functions (Kuzala & Vargo, 1991) Measurement of grip strength is a basis to use in the physical medicine to assess patients work capacities and progress of rehabilitation (Gilbert & Knowlton, 1983). Functional Hand evaluation is a kind of performance test, its validity, therefore relies heavily on the cooperation or sincerity of the subjects being assessed.

Previous studies of Grip sincerity Static Isometric Grip Test Approach (Stokes 1983, Stoke et. al 1995, Niebuhr and Marion 1987, Niebuhr and Marion 1990) Rapid Alternate Grip Test Approach (Hildreth, Breidenbach, Lister and Hodges 1989, Joughin et al 1993) Coefficient of Variation Approach (Bechtol 1954, Robinson et al 1993, Fairfax, Balnave & Adams 1995) Sustain Grip Test Approach (Kroemer & Marras 1980, Gilbert and Knowlton 1983, Smith et al 1989, Chengalur et al 1990)

Chenglaurs Method (1990) The EVAL system by Greenleaf Medical Systems was used in this research Typical Force-Time Curve For Sincere & Fake Condition Time (Sec) Grip Strength (Lbs) Fake Trial Sincere Trial P=Peak force A=average of plateau SD=SD of plateau

The discriminators (Chengalur,1990) Derived from data obtained from the sustain grip test D1. Ratio: 100 * A / P (greater – more sincere) D2. Coefficient of Variation: 100 * SD of Plateau / Mean of Plateau (smaller – more sincere) D3 called Ratio Difference. It was devised for comparison between major & minor hand (Maj / Min) characteristics. (for healthy subjects only: greater – more sincere) D4: Peak-Average Difference: {(P-A) * 100 }/ (P *A) (smaller – more sincere) D5: Peak-Average Root Difference : (smaller – more sincere)

Criterion value (method proposed by Chengalur,1990) *Remarks: Stand for the cut-off point using 95% confidence level (If subject has D4 score equal or less than the cut-off point, it is 95% sure the subject lies within the sincere group) D4 Score for Sincere And Fake Trial – hypothetical (smaller – more sincere) D4 Score Frequency Sincere Trial Fake Trial Zone I Zone II cut-off point

Objectives of current study To test the applicability of Chengalurs methods and its findings in Chinese subjects. To explore the use of multivariate analysis (logistic regression) on the raw data obtained in the sustain grip test to detect faking.

Demographic Profile of Healthy Subjects (Group 1)

Demographic Profile of Ex-hand Injured Patients (Group 2)

Test Procedure All subjects are taught to do the sustain grip test using the EVAL System. All subjects were randomized into 2 subgroups – subgroup 1a perform sincere grip and then fake grip, while subgroup 1b perform fake grips first and then sincere grip, etc. For fake grips, subjects are assigned randomly a faking ratio of 25%, 50%, or 75% of maximum grip strength.

Test Procedure Findings from sustain grip tests were converted into the discriminator, i.e. D1, D2, D4, D5. (D3 was used to healthy subjects only). % of accurate detection of sincere and fake group by using the Chengulars method and Logistic regression was estimated and compared.

Instrumentation The EVAL system by Greenleaf Medical Systems was used in this research Typical Force-Time Curve For Sincere & Fake Condition Time (Sec) Grip Strength (Lbs) Fake Trial Sincere Trial P=Peak force A=average of plateau SD=SD of plateau

Hypothetical distribution of D4 showing the application of criterion value in determining fake and sincere group Remarks: the cut-off point Zone I:Under the blue curve - true sincere grip (87.5%) Under the red curve - false sincere grip (type II error, 21.2%) Zone II:Under the blue curve - false fake grip (type I error, 12.5%) Under the red curve - true fake grip (77.8%) Hypothetical distribution of D4 (male) 0 D4 Value Frequency Sincere Grip Fake Grip Zone I Zone II 1.10

Percentage of successful detection of fake grip by Discriminant Analysis (Group 1-healthy subjects)

Percentage of successful detection of fake grip by Logistic Regression (Group 1-healthy subjects)

Percentage of successful detection of fake grip by Discriminant Analysis (Group 2-Ex-patients)

Percentage of successful detection of fake grip by Logistic Regression (Group 2-Expatients)

Percentage of successful detection of fake grip by Discriminant Analysis (Groups 1 & 2)

Percentage of successful detection of fake grip by Logistic Regression (Groups 1 & 2)

Results-Summary Both methods, i.e. Discriminant Analysis and Logistic Regression can detect between 90%- 100% sincere grips. For subjects, who performed at 20% or 50% of their maximum grips, it was possible to detect 92% faking in males and 86.5% in female by using logistic regression method. True sincere True faking

Conclusions It is possible to obtain better prediction accuracy by using the multivariate statistical method of approach (Logistic regression). It is also possible to formulate a mathematical model for detecting the faking grip. This, however will depend on a larger sample for developing a more reliable formula for clinical and medico-legal application.