Academy of Management, New Orleans, 2004 1 Taking a crack at measuring faultlines Sherry M.B. Thatcher (University of Arizona) Katerina Bezrukova (Rutgers.

Slides:



Advertisements
Similar presentations
Introduction Simple Random Sampling Stratified Random Sampling
Advertisements

Clustering Clustering of data is a method by which large sets of data is grouped into clusters of smaller sets of similar data. The example below demonstrates.
© McGraw-Hill Higher Education. All rights reserved. Chapter 3 Reliability and Objectivity.
Marketing Research Aaker, Kumar, Day and Leone Tenth Edition Instructor’s Presentation Slides 1.
K Means Clustering , Nearest Cluster and Gaussian Mixture
Advanced Methods and Models in Behavioral Research – 2014 Been there / done that: Stata Logistic regression (……) Conjoint analysis Coming up: Multi-level.
Jarvenpaa, CORE 12/15/02 Geographical Diversity in Global Virtual Teams Jeffrey T. PolzerC. Brad Crisp Harvard UniversityIndiana University Sirkka L. JarvenpaaWon-Yong.
AEB 37 / AE 802 Marketing Research Methods Week 7
Statistics for Marketing & Consumer Research Copyright © Mario Mazzocchi 1 Cluster Analysis (from Chapter 12)
I OWA S TATE U NIVERSITY Department of Animal Science Model Development and Selection of Variables Animal Science 500 Lecture No. 17 October 28, 2010.
1 Single Indicator & Composite Measures UAPP 702: Research Design for Urban & Public Policy Based on notes by Steven W. Peuquet. Ph.D.
Review: What influences confidence intervals?
© 2005 The McGraw-Hill Companies, Inc., All Rights Reserved. Chapter 14 Using Multivariate Design and Analysis.
Nemours Biomedical Research Statistics April 2, 2009 Tim Bunnell, Ph.D. & Jobayer Hossain, Ph.D. Nemours Bioinformatics Core Facility.
Cluster Analysis (1).
Longitudinal Data Analysis: Why and How to Do it With Multi-Level Modeling (MLM)? Oi-man Kwok Texas A & M University.
Reliability of Selection Measures. Reliability Defined The degree of dependability, consistency, or stability of scores on measures used in selection.
LEARNING PROGRAMME Hypothesis testing Intermediate Training in Quantitative Analysis Bangkok November 2007.
Fundamentals of Statistical Analysis DR. SUREJ P JOHN.
hss2381A – quantitative methods
Instrumentation.
Moderators. Definition Moderator - A third variable that conditions the relations of two other variables Example: SAT-Quant and math grades in school.
Descriptive Statistics
Consistency Matters! The Multilevel Effects of Group and Division Cultures on the Faultline-Outcomes Link Katerina Bezrukova ( Rutgers University) Sherry.
VI. Evaluate Model Fit Basic questions that modelers must address are: How well does the model fit the data? Do changes to a model, such as reparameterization,
Cluster analysis 포항공과대학교 산업공학과 확률통계연구실 이 재 현. POSTECH IE PASTACLUSTER ANALYSIS Definition Cluster analysis is a technigue used for combining observations.
Multilevel Data in Outcomes Research Types of multilevel data common in outcomes research Random versus fixed effects Statistical Model Choices “Shrinkage.
Paper: Large-Scale Clustering of cDNA-Fingerprinting Data Presented by: Srilatha Bhuvanapalli INFS 795 – Special Topics in Data Mining.
Univariate modeling Sarah Medland. Starting at the beginning… Data preparation – The algebra style used in Mx expects 1 line per case/family – (Almost)
1 of 14 Exploring race and gender differentials in student ratings of instructors: Lessons from a diverse liberal arts college Robert L. Moore, Hanna Song.
Hypothesis testing Intermediate Food Security Analysis Training Rome, July 2010.
Handbook on Residential Property Price Indices Chapter 5: Methods Jan de Haan UNECE/ILO Meeting, May 2010.
PHP Form Introduction Getting User Information Text Input.
5-4-1 Unit 4: Sampling approaches After completing this unit you should be able to: Outline the purpose of sampling Understand key theoretical.
Susan O’Shea The Mitchell Centre for Social Network Analysis CCSR/Social Statistics, University of Manchester
Intermediate Applied Statistics STAT 460 Lecture 17, 11/10/2004 Instructor: Aleksandra (Seša) Slavković TA: Wang Yu
Advanced Meta-Analyses Heterogeneity Analyses Fixed & Random Efffects models Single-variable Fixed Effects Model – “Q” Wilson Macro for Fixed Efffects.
Session 3, Unit 5 Dispersion Modeling. The Box Model Description and assumption Box model For line source with line strength of Q L Example.
U Describes the relationship between two or more variables. Describes the strength of the relationship in terms of a number from -1.0 to Describes.
Right Hand Side (Independent) Variables Ciaran S. Phibbs June 6, 2012.
HLM Models. General Analysis Strategy Baseline Model - No Predictors Model 1- Level 1 Predictors Model 2 – Level 2 Predictors of Group Mean Model 3 –
Understanding Your Data Set Statistics are used to describe data sets Gives us a metric in place of a graph What are some types of statistics used to describe.
Math 341 January 23, Outline 1. Recap 2. Other Sampling Designs 3. Graphical methods.
ICCS 2009 IDB Workshop, 18 th February 2010, Madrid 1 Training Workshop on the ICCS 2009 database Weighting and Variance Estimation picture.
Right Hand Side (Independent) Variables Ciaran S. Phibbs.
Review I A student researcher obtains a random sample of UMD students and finds that 55% report using an illegally obtained stimulant to study in the past.
Personal Control over Development: Effects on the Perception and Emotional Evaluation of Personal Development in Adulthood.
Applied Multivariate Statistics Cluster Analysis Fall 2015 Week 9.
Mx modeling of methylation data: twin correlations [means, SD, correlation] ACE / ADE latent factor model regression [sex and age] genetic association.
The Mixed Effects Model - Introduction In many situations, one of the factors of interest will have its levels chosen because they are of specific interest.
Two-Factor Study with Random Effects In some experiments the levels of both factors A & B are chosen at random from a larger set of possible factor levels.
Quantitative methods and R – (2) LING115 December 2, 2009.
Math 145 June 19, Outline 1. Recap 2. Sampling Designs 3. Graphical methods.
Carina Omoeva, FHI 360 Wael Moussa, FHI 360
Testing for moderators
School Quality and the Black-White Achievement Gap
Math 145 January 23, 2007.
Standard Deviation.
Advanced Quantitative Analysis
Gerald - P&R Chapter 7 (to 217) and TEXT Chapters 15 & 16
Soc 3306a: ANOVA and Regression Models
A New Approach to the Study of Teams: The GAPIM
LAMAS Working Group 7-8 December 2015
Neural Networks and Their Application in the Fields of Coporate Finance By Eric Séverin Hanna Viinikainen.
Soc 3306a Lecture 11: Multivariate 4
Cluster analysis Presented by Dr.Chayada Bhadrakom
Exercise 1 Use Transform  Compute variable to calculate weight lost by each person Calculate the overall mean weight lost Calculate the means and standard.
Graziano and Raulin Research Methods: Chapter 12
Presentation transcript:

Academy of Management, New Orleans, Taking a crack at measuring faultlines Sherry M.B. Thatcher (University of Arizona) Katerina Bezrukova (Rutgers University) Karen A. Jehn (Leiden University)

Academy of Management, New Orleans, Agenda Interactive Exercise Why? –Importance of faultlines vs. other composition measures How? –What we did Huh? –Problems we ran into (and how we fixed them) Oh, that! –Issues that journal reviewers are likely to raise

Academy of Management, New Orleans, Interactive exercise

Academy of Management, New Orleans, Interactive exercise In breaking the group into subgroups, what characteristics did you look at? How homogeneous are the subgroups? What assumptions did you make when breaking the group into subgroups?

Academy of Management, New Orleans, Why? Mixed effects of diversity and demography studies Focus on more than one attribute at a time Takes into account interdependence among attributes

Academy of Management, New Orleans, How? From Diversity to Faultlines Step 1: Picturing what we need to measure ♀♀P♀♀P ♀♀P♀♀P ♀♀P♀♀P ♀♀P♀♀P ♀♀P♀♀P ♀♀P♀♀P Educ. Race Sex ♂♂P♂♂P ♀♀H♀♀H ♂♂P♂♂P ♀♀H♀♀H ♂♂P♂♂P ♀♀H♀♀H Educ. Race Sex Group A: Strong FaultlineGroup B: Weak Faultlines  H = High school, P = PhD, W = White, B = Black, M = Male, F = Female HWM PBF HWM HBF PBM PWF

Academy of Management, New Orleans, How? Step 2: Understanding diversity formulas [1/n  (X i - X j ) 2 ] 1/2 ] Individual-level categorical and interval variables. Relational demography /individual dissimilarity score (Tsui & O’Reilly, 1989). SD Group-level interval variables. Coefficient of variation (Allison, 1978). (1 –  Pi   Group-level categorical variables. Index of heterogeneity (Blau, 1977; Bantel & Jackson, 1989); Diversity or entropy index (Teachman, 1980; Ancona & Caldwell, 1992). x

Academy of Management, New Orleans, How? Step 3:Creating a faultline strength formula Faultline strength – Clustering Algorithm based on Euclidean distance formula (Thatcher, Jehn, & Zanutto, 2003) –x ijk = the value of the j th characteristic of the i th member of subgroup k –x j = the overall group mean of characteristic j –x jk = the mean of characteristic j in subgroup k –n g k = the number of members of the k th subgroup (k=1,2) under split g –the faultline strength = the maximum value of Fau g over all possible splits g=1,2,…S.

Academy of Management, New Orleans, Measuring Faultlines (strongest split is AC, BD but AB, CD is also a strong split) Weak (1 align; 4 ways) (strongest split is AB, CD) Very Strong (4 align; 1 way) (strongest split is AC, BD) Strong (3 align; 2 ways) ( strongest split is AB, CD, but BC, AD is also close ) Weak (1 align; 3 ways) 0None FAU ALGORITHM based on Euclidean distance formula FAULTLINE STRENGTH/ L & M

Academy of Management, New Orleans, How? Revisiting Step 1: Faultline Distance Faultline distance reflects how far apart the subgroups are from each other Age Education Tenure Age Education Tenure 3055 M.S.Ph.D Ph.D B.A. 3  Group B: Closer TogetherGroup A: Farther Apart

Academy of Management, New Orleans, Faultline Distance (cont’d) Faultline distance - the Euclidean distance between the two sets of averages where centroid (vector of means of each variable) for subgroup 1 = ( ), centroid for subgroup 2 = ( ). Group faultline score Fau = Strength (Fau g ) x Distance (D g )

Academy of Management, New Orleans, Faultlines Strength and Distance, and Group Faultlines Scores MemberAgeRaceGenderTenureFunctionEducation Team Team Faultline Strength Faultline Distance Group Faultlines Score

Academy of Management, New Orleans,

Academy of Management, New Orleans,

Academy of Management, New Orleans, Rescaling Considerations Theory driven approach –to use SME’s judgments to weight characteristics Empirical approach –to view participants’ responses as a “true” measure of faultlines Statistical approach –to use standard deviations

Academy of Management, New Orleans,

Academy of Management, New Orleans,

Academy of Management, New Orleans,

Academy of Management, New Orleans,

Academy of Management, New Orleans,

Academy of Management, New Orleans,

Academy of Management, New Orleans, SAS Faultline Calculation (Version 1.0, July 26, 2004) 1.WHAT THIS CODE DOES faultline strength and distance for groups of size 3 to 16 (two sets: incl and excl 1-person subgroups). 2.WHAT WE ASSUME ABOUT THE DATA a comma-separated data text file (save as.csv file). dummy variables for categorical vars. no missing values group ID variable (groups are numbered from 1 to n) 3.WHAT WE ASSUME ABOUT THE RESCALING FACTORS rescaling factors must be specified for each variable rescaling factors must be specified in a comma-separated text file (save as.csv file).

Academy of Management, New Orleans, SAS Faultline Calculation (Version 1.0, July 26, 2004): Cont’d 4. HOW TO RUN THE CODE –download the SAS code and data files into C:\Faultline\FL_code\FL_Code_parameters.txt –go to the C:\Faultline\FL_Code directory and double click on FL_Code_1_0.sas –right click the mouse and select “Submit All” 5. HOW TO MODIFY THE INPUT PARAMETERS –all user inputs are specified in the file C:\Faultline\FL_Code\FL_Code_parameters.txt. –keep exact names of files.

Academy of Management, New Orleans, Huh? Problems we ran into (and how we fixed them) Group size Number of possible subgroups Subgroups of size “1” Calculating the overall faultline score Measuring faultline distance for categorical variables Rescaling

Academy of Management, New Orleans, Oh That! Issues that journal reviewers have raised Rescaling (influence on results) –solution: rerun analyses Importance of distance component –solution: explain it better Perceptual faultlines = actual faultlines? –solution: explain to the reviewers that we didn’t have this data

Academy of Management, New Orleans, Advantages of Fau Measure allows continuous and categorical variables unlimited number of variables theoretically unlimited group size flexible enough to allow for different rescaling

Academy of Management, New Orleans, Future Research & Work in Progress Testing the theory in experimental settings Faultlines, coalitions, conflict, group identity and leadership profiles Temporal effects of faultlines Testing the theory in organizational settings Consistency matters! The Effects of Group and Organizational Culture on the Faultline-Outcomes Link Testing the theory in international settings Peacekeeping and Ethnopolitical conflict A quasi-experimental field study in ethnic conflict zones (i.e., Crimea, Sri Lanka, Burundi and Bosnia)

Academy of Management, New Orleans, Thank you very much for coming Any questions?