what is multidimensional scaling (uw)

Upload: rahul-mishra

Post on 09-Apr-2018

236 views

Category:

Documents


0 download

TRANSCRIPT

  • 8/8/2019 What is Multidimensional Scaling (UW)

    1/17

    Mod: 11/ 2009

    What is MDS?

    Prof APM Coxon, U Cardiff 1

    What isWhat is

    MultidimensionalMultidimensional

    Scaling [MDS] ?Scaling [MDS] ? Anthony P.M. Coxon

    Emeritus Professor of Sociological Research Methods, Universityof Wales

    Honorary Professor, Cardiff University

    Honorary Professorial Research Fellow, University of Edinburgh Co-founder & Co- Director of MDS software packages,

    MDSX [OS] (freeware)and

    NewMDSX for Windows (not-for-profit)

    Website:www.newmdsx.com

    Course materials: http://apmc.newmdsx.com/

    see my entry on multidimensional scaling inLewis-Beck, M.S. et al, eds (2004) The Sage Encyclopaedia ofSocial Science Research Methods. London Sage Publications )

    http://www.newmdsx.com/http://apmc.newmdsx.com/http://apmc.newmdsx.com/http://www.newmdsx.com/
  • 8/8/2019 What is Multidimensional Scaling (UW)

    2/17

    U Winchester 12/09 What is MDS?Prof APM Coxon, Cardiff U 2

    ORIGINS / DEVELOPMENT OF MDSORIGINS / DEVELOPMENT OF MDS

    MDS (aka Smallest Space Analysis) Has origins in Psychometrics in 1920-60s: Scale construction and dimensionality reduction

    Underwent major burst of development in 1960s due tonon-metric revolution(Coombs) and computingdevelopments allowing iterative estimation

    Originally designed for analysis of LTM of dis/similarities data ,taking a range of measures (not just PM correlations):

    anything which, by an act of faith, can be considered a similarity(Shepard)

    Extended rapidly to deal with wide range of other types of data Rectangular matrices ; triads, pair-comparisons, free-sorting

    stacks of matrices (3-way scaling INDSCAL)

  • 8/8/2019 What is Multidimensional Scaling (UW)

    3/17

    3

    CONSTRUCTING A MAP

    Given a map, its easy to calculate the distances between

    the points MDS operates the other way round:

    Given the data [ interpreted as quasi distances ] itattempts to find the configuration [location ofpoints] which generated the distances

    This is Classic MDS: developed in 1930s butimperfect, not robust, & works only if data are ratio.

    Whereas more recent MDS can work when only theordinal information exists: Non-metric = ordinalMDS (Coombs / Kruskal non-metric revolution)

    What?? You can create an accurate map from onlythe rankorder of the distances???

    Yes! And it works!!

  • 8/8/2019 What is Multidimensional Scaling (UW)

    4/17

    The RANK of distances can recover the Mapthough not the coastline

    NEWMDSX(RUNSCRIPT + SYNTAX)

    RUN NAME Rank of Scottish distances,

    COMMENT 1 = smallest; 120 = max; dissimilarity data

    F3.3, p48 The Users Guide to MDS

    N OF STIMULI 16

    PARAMETERS DATA TYPE(1)

    LABELS BERWICK

    EDINBURGH

    GLASGOW

    STRANRAER

    AYR

    PERTH

    DUNDEE

    ABERDEEN

    STIRLING

    OBANFORT_WM

    INVERNESS

    KYLE_LOCHALSH

    BRAEMAR

    ULLAPOOL

    THURSO

    READ MATRIX

    17

    53 11

    92 68 36

    70 30 4 11

    34 7 19 83 45

    27 8 29 93 58 1= Perth-Dundee63 56 83 115 103 35 24

    43 4 2 58 21 3 14 66

    99 57 26 72 36 43 63 98 28

    96 60 39 89 62 41 49 79 33 6

    100 75 75 112 97 45 49 48 60 52 22

    111 89 78 107 89 72 81 94 70 26 9 23

    67 36 51 106 80 15 10 19 30 65 30 15 54

    114 105 101 119 109 85 87 86 88 68 40 16 17 55

    117 113 115 120 119 103 102 77 110 107 95 47 62 74 42

    COMPUTE

    = Stranraer - Thurso

    U. Winchester12/09

    What is MDS?Prof APM Coxon, Cardiff U

    4

  • 8/8/2019 What is Multidimensional Scaling (UW)

    5/17

    Uni Winchester 12/09

    What is MDS?

    Prof APM Coxon, Cardiff Uni 5

    WHAT IS MULTIDIMENSIONALSCALING?

    A students definition: If you are interested in how certain objects relate to each

    other and if you would like to present these relationshipsin the form of a map then MDS is the technique you need

    (Mr Gawels, KUB)A good start!

    MDS provides a useful and easily-assimilable graphic visualisation of all

    sorts of data

    Tukey: A picture is worth a thousand words

    In a user-chosen (small) # of dimensions

    providing a graphical representation of the structure

    underlying a complex data set And measure how well / badly the solution distances matchthe data dissimilarities (Stress)

  • 8/8/2019 What is Multidimensional Scaling (UW)

    6/17

    U Winchester, 12/2009

    What is MDS?

    Prof APM Coxon, Cardiff U 6

    MDS is a family of modelsdifferentiated by

    (DATA) the empirical inter-relationships between

    a set of objects/variables which are given in aset of dis/similarity data

    Basically, type of input data, defined by their Way andMode [e.g. 2W1M]. (Cf observations vs data)

    (FUNCTION) data are then optimally re-scaled

    (according to permissible trans-formations for thedata) in terms of

    Choice of level of measurement [e.g. ordinal ]

    (MODEL) the assumptions of the model chosen torepresent the data

    Usually (Euclidean) Distance model

  • 8/8/2019 What is Multidimensional Scaling (UW)

    7/17

    Uni Winchester 12/09 What is MDS?Prof APM Coxon, Cardiff Uni

    7

    data

    MDS can be used with a wide variety ofDATA

    e.g.: SORTS OF DATA direct data (pair comparisons, ratings, rankings,

    triads, counts)

    derived data (profiles, co-occurrence matrices,textual data, aggregated data)

    measures of association etc derived from simpler

    data, and

    tables of data.

    TYPES of DATA

    Described by WAY (2W=matrix; 3W=stack of matrices )

    And MODE (# sets of distinct objects eg variables,subjects) E.G. 2W1M; 2W2M; 3W2M 7W4M

  • 8/8/2019 What is Multidimensional Scaling (UW)

    8/17

    Uni Winchester 12/09

    What is MDS?

    Prof APM Coxon, Cardiff U 8

    VARIANTS OF MDS MODELS due toTRANSFORMATIONS

    MDS can also be used with a wide variety of:

    Transformations (levels of measurement) monotonic (ordinal),

    linear/metric (interval),

    but also

    Splines (SPSS PROXSCAL) local preservation of distance

    log-interval (MRSCAL),

    Power (MULTISCALE)

    smoothness

  • 8/8/2019 What is Multidimensional Scaling (UW)

    9/17

    VARIANTS OF MDS due to type of MODEL

    DISTANCE Minkowski-r Usually Euclidean (r=2)

    Less often City Block, r=1

    Sometimes Dominance,r= 32

    SCALAR PRODUCTS/Factor scalar product : a b = |a| |b| cos

    E.g. Covariance, PM Correlation

    As used in PCA, FA, MDPREF

    COMPOSITION Most usually, Additive (cf ANOVA), as in

    Impression Formation:

    X(i.j) = a(i) + b(j) + nb Ordinal.non-metric ANOVA

    But also, difference, product, mixed

    Uni Winchester 12/09

    What is MDS?

    Prof APM Coxon, Cardiff Uni 9

  • 8/8/2019 What is Multidimensional Scaling (UW)

    10/17

    HOW DOES MDS WORK? Iteratively!

    START: Produce Init. Guestimate Configuration (a) FIT

    Calculate distances (d)

    Compare with data () [via Ordinal regression]

    Calculate overall badness -of-fit measure

    Stress (d- ) well, almost! Actually more complex

    Perfect/Acceptable? EXIT

    (b) IMPROVE: For each point, find direction of improvement (dont ask: calculus! Derivatives!)

    How far to move? Step-size (call it heuristic ; parachute & mist)

    (c) MOVE configuration/points

    BACK TO (a)

    Uni Winchester 12/09

    What is MDS?

    Prof APM Coxon, Cardiff Uni 10

  • 8/8/2019 What is Multidimensional Scaling (UW)

    11/17

    Uni Winchester 12/2009

    What is MDS?

    Prof APM Coxon, Cardiff Uni 11

    MDS PROGRAMS:1. Usually either General Purpose Package (SPSS )

    Basic Model for 2W1M data: PROXSCAL and 3W2M INDSCAL Also contains CORRESP, HICLUS and (in >SPSS13 ) PREFSCAL (2W2M)

    2. or Library : set of programs, each specific to Data-shape, Trans & Model (e.g. NewMDSX for Windows); includes

    BASIC 2W1M SCALING:

    Non-metric (ordinal) MINISSA , Metric (MRSCAL) linear,

    Clustering (Hierarchical & Non-hierarchical)

    2W2M (Rectangular) SCALING:

    Multidimensional Preference, Triads, Unfolding, Sorting

    3W2M (and higher) SCALING: Individual Differences (INDSCAL), (Tucker) Points-of-View

    Procrustean IndDiffs (Lingoes PINDIS)

    Or Interactive Package (PERMAP via NewMDSX)

    primarily for basic model

    Visually animated

    Superb diagnostic procedures

  • 8/8/2019 What is Multidimensional Scaling (UW)

    12/17

    Uni Winchester, 12/09

    What is MDS?

    Prof APM Coxon, Cardiff Uni 12

    SITES & SOFTWARE:

    SITES NEWMDSX AND DOCUMENTATION:

    http://www.newmdsx.com

    INTERACTIVE PERMAP (Heady)

    (presently obtained via NewMDSX)

    THREE-WAY SCALING (Kroonenberg)

    http://www.leidenuniv.nl/fsw/three-mode/content.htm

    FORREST YOUNGS VISTA (Visual Statistics)http://forrest.psych.unc.edu/research/index.html

  • 8/8/2019 What is Multidimensional Scaling (UW)

    13/17

    UniWinchester 12/09

    What is MDS?

    Prof APM Coxon, Cardiff Uni 13

    WHAT IS MDS?

    and now for anexample!

  • 8/8/2019 What is Multidimensional Scaling (UW)

    14/17

    APPENDICES

    1. Interpretation: Headlines

    2. MVA & MDS

    Professor APM Coxon 14

  • 8/8/2019 What is Multidimensional Scaling (UW)

    15/17

    MDS: Interpretation: Headlines

    For Euclidean Distance MDS: "What information isstable/significant?

    Beware Local minima [PERMAP]

    Remember: You may translate, reflect, (rigidly) rotatethe configuration: do so! [e.g. NewMDSX Graphics;PERMAP]

    CLEARING UP Configuration: [PERMAP] Map Evaluation & Diagnostics; Points and Links;

    selective removal and hints of structure via WaernsGraphic links.

    BASIC STRUCTURES: Regional: what points are close to each other and distant

    from others? CLUSTERING [(HI)CLUS, SPSS]

    Linear: directions in space where some property isincreasing: External properties [PRO-FIT NewMDSX],

    If you must ... dimensions -- remember changing theorigin or dimensional orientation has no effect onrelative distance. Most MDS rotated at end to PCA ...Unlike FA, dimensions may/ may not have importance.

    SIMPLE STRUCTURES dimensions, yes -- but also other simple structures

    (horseshoes, radex/circumplex).Professor APM Coxon 15

  • 8/8/2019 What is Multidimensional Scaling (UW)

    16/17

    MDS & other DimensionalMultivariate Analysis models

    Uni Winchester 12/09

    What is MDS?

    Prof APM Coxon, Cardiff Uni 16

  • 8/8/2019 What is Multidimensional Scaling (UW)

    17/17

    Research MethodsFestival 2006

    What is MDS?

    Prof APM Coxon, U Edinburgh 17

    SOME POSSIBLE WEAKNESSES inMDS

    There ARE any??! Relative ignorance of the sampling/inferential properties ofstress But, simulation (Spence), MLE estimation

    Prone-ness to local minima solutions but less so, and multiple starts & interactive programs like PERMAP allow

    thousands of runs to check

    A few forms of data/models are prone to degeneracies especially MD Unfolding, but see new PREFSCAL in SPSS14)

    difficulty in representing the asymmetry of causal models though external analysis is very akin to dependent-independent

    modelling,

    there are convergences with GLM in hybrid models such as CLASCAL

    (INDSCAL with parameterization of latent classes)