1. Validation o f Order Rank Scales Based o n Compositional Data Analysis: A Proposal. Claudia Malpica Lander [email_address] Purificación Galindo Villardón [email_address] Universidad de Salamanca Departamento de Estadística
2. INTRODUCTION Validity Description and of order rank scales. Compositional data (AITCHISON, 1986) An application is presented in the management evaluation context.
3. LEADERSHIP DIMENSIONS BOLMAN AND DEAL (1991) STRUCTURAL HUMAN RESOURCE POLITICAL SYMBOLIC Roots: personality and social psychology Key concepts: needs (motives), capacities (skills), feelings Central focus: fit between individual and organization Roots: sociology, management science Key concepts: goals, roles (division of labor), formal relationships Central focus: alignment of structure with goals and environment Roots: political science Key concepts: interests, conflict, power, scarce resources Central focus: getting and using power, managing conflict to get things done Roots: social and cultural anthropology Key concepts: culture, myth, ritual, story, Central focus: building culture, staging organizational drama
4. CONSTRUCT VALIDITY OF LIKERT´S SCALE Validity 1.- Think very clearly and logically. 4 2.- Show high levels of support and concern for others. 5 3.- Have exceptional ability to mobilize people. 3 4.- Inspire others to do their best. 4 5.- Strongly emphasize careful planning and clear times lines. 4 … … … … … 31.- Succeed in the face of conflict and opposition. 5 32.- Serve as an influential model of organization aspirations and values. 3 1 = Never 2 = Occasionally 3 = Sometimes 4 = Often 5 = Always Exploratory and Confirmatory Factor Analysis 29 e29 1 1 25 e25 1 21 e21 1 17 e17 1 13 e13 1 i9 e9 1 5 e5 1 1 e1 1 29 e29 1 1 25 e25 1 21 e21 1 17 e17 1 13 e13 1 e9 1 5 e5 1 1 e1 1 i 30 e 30 1 1 2 6 e2 6 1 2 2 e2 2 1 1 8 e1 8 1 1 4 e1 4 1 i 10 e 10 1 i 6 e 6 1 i 2 e2 1 i 30 e 30 1 1 2 6 e2 6 1 2 2 e2 2 1 1 8 e1 8 1 1 4 e1 4 1 i 10 e 10 1 i 6 e 6 1 i 2 e2 1 i 31 e 31 1 1 2 7 e2 7 1 2 3 e2 3 1 1 9 e1 9 1 1 5 e15 1 i 11 e 11 1 i 7 e 7 1 i 3 e3 1 i 31 e 31 1 1 2 7 e2 7 1 2 3 e2 3 1 1 9 e1 9 1 1 5 e15 1 i 11 e 11 1 i 7 e 7 1 i 3 e3 1 i 31 e 31 1 1 2 7 e2 7 1 2 3 e2 3 1 1 9 e1 9 1 1 5 e15 1 i 11 e 11 1 i 7 e 7 1 i 3 e3 1 i 31 e 31 1 1 2 7 e2 7 1 2 3 e2 3 1 1 9 e1 9 1 1 5 e15 1 i 11 e 11 1 i 7 e 7 1 i 3 e3 1 STRUCTURAL HHRR POLITICAL SYMBOLIC
5. CONSTRUCT VALIDITY OF ORDER RANK SCALE Validity 1.- What has helped me the most to be successful is my ability to: ____A. Make good decisions. ____B. Coach and develop people. ____C. Build strong alliances. ____D. Energize and inspire others . … … … … … 6.- The best way to describe me is: ____A. Technical expert. ____B. Good listener. ____C. Skilled negotiator. ____D. Inspirational Leader . 2 3 1 4 4 1 2 3 4 = best describes 3 = next best … 1 = least like you Every question can be considered as a composition with total equal to 10 Validity A = STRUCTURAL B = HHRR C = POLITICAL D = SYMBOLIC Every dimension performes Laedership composition adding 60 Description
7. COMPOSITIONAL DATA Validity Self Supervisors Peers Subordinates Ind1 Ind2 Ind3 Indn 1A 1B 1C . . . 6D BIPLOTS OF COMPOSITIONAL DATA Z=U V T (GREENACRE & AITCHISON, 2002) Z R D DOUBLE CENTRED l = log(x) x S D x S D AITCHISON´S DISTANCE (1992) CENTRED LOGRATIO, CLR Z R D g(x) x log clr(x) z (AITCHISON, 1986)
10. STATIS DUAL Validity Z 1 Z 2 Z K J I 1 I T I 2 INTER-STRUCTURE ANALYSIS COMPROMISE STRUCTURE INTRA-STRUCTURE ANALYSIS Z R D DOUBLE CENTRED x S D STATIS DUAL (LAVIT, 1988) S tructuration de T ableaux A T rois I ndices de la S tatistique (GREENACRE & AITCHISON, 2002)
11. STATIS DUAL Validity J J COMPROMISE STRUCTURE C 1 C 2 C k J J J J RV matrix K K C 2 C 4 C 1 C 3 INTER-STRUCTURE ANALYSIS PCA PCA J2_1 J1_1 J3_1 J1 _2 J2_2 J3_2 J2_3 J1_3 J3_3 J2_4 J1_4 J3_4 INTRA-STRUCTURE ANALYSIS J2_1 J1 J 2 J3 J4 J5 J6 J7 J8 J9 J1 0 J11
17. CONSIDERING EXTERNAL INFORMATION Description Ind1 Ind2 Ind3 Indn SEX AGE . . . Y K K MATRIX CATEGORIES MALE FEMALE . . . Self (n = 1245) Supervisors (n = 1444) Peers (n = 2343) Subordinates (n = 2448) Ind1 Ind2 Indn STRU HHRR POLI SYMB Z K DOUBLE CENTRED STRU HHRR POLI SYMB
18. MATRICES W K Description Male Female . . . 25-34 STRU HHRR POLI SYMB W k Self Supervisors Peers Subordinates STATIS Dual Z k Ind1 Ind2 Ind3 Indn STRU HHRR POLI SYMB Mas Fem . . . Ind1 Ind2 Ind3 Indn 25-34 Where = diag (y 1· ) Y k W k STRU HHRR POLI SYMB Male Female . . . 25-34 Canonical Correspondence Analysis (TER BRAAK, 1986)
19. STATIS DUAL W K Description -0.7 -0.6 -0.5 -0.4 -0.3 -0.2 -0.1 0 0.1 0.2 0.3 -0.5 -0.4 -0.3 -0.2 -0.1 0 0.1 0.2 0.3 Male S Fema S 2534 S 3544 S 4554 S 5564 S tecn S lice S maes S prof S g-su S g-me S hi-g S Male U Fema U 2534 U 3544 U 4554 U 5564 U tecn U lice U maes U prof U g-su U g-me U hi-g U Male P Fema P 2534 P 3544 P 4554 P 5564 P tecn P lice P maes P prof P g-su P g-me P hi-g P Male B Fema B 2534 B 3544 B 4554 B 5564 B tecn B lice B maes B prof B g-su B g-me B hi-g B STRU HHRR POLI SYMB Structural Political Symbolic HHRR
20. STATIS DUAL W K Description -0.5 -0.4 -0.3 -0.2 -0.1 0 0.1 0.2 0.3 -0.7 -0.6 -0.5 -0.4 -0.3 -0.2 -0.1 0 0.1 0.2 0.3 Political Symbolic Structural HHRR S S S S S S S S S S S U U U U U U U U U P P P P P P P P B B B B B B B S U U U U P P B B B B S P P B B S elf (n = 1245) S U pervisors (n = 1444) P eers (n = 2343) Su B ordinates (n = 2448)
21. STATIS DUAL W K Description -0.5 -0.4 -0.3 -0.2 -0.1 0 0.1 0.2 0.3 -0.7 -0.6 -0.5 -0.4 -0.3 -0.2 -0.1 0 0.1 0.2 0.3 Political Symbolic Male _S Fema _S Male _U Fema _U Male _P Fema _P Male _B Fema _B Structural HHRR S elf (n = 1245) S U pervisors (n = 1444) P eers (n = 2343) Su B ordinates (n = 2448) Male Fema le Sexo