March 14, 2005Methodology & Statistics ANOVA (GLM- Repeated measures) Jantien Donkers

March 14, 2005Methodology & Statistics When to use.. oT-test: 2 conditions testing 1 independent variable (e.g. text complexity – simple/complex in relation to number of recalled words) oANOVA: 3 or more conditions testing 1 (one way ANOVA) or more (two way ANOVA) variables

March 14, 2005Methodology & Statistics ANOVA’s oF-ratio oSize of the variance due to the experimental conditions in relation to the error (unexpected) variance oDegrees of freedom (df)

March 14, 2005Methodology & Statistics ANOVA’s (2) oH0: all means are equal oAlternative: all of means are different, or just one of them oVariation among groups is compared to variation within groups: a relatively large difference is evidence against H0

March 14, 2005Methodology & Statistics Parametric assumptions oExperimental scores are measured on an interval scale oScores are normally distributed oVariabiliy of scores for each conditions should be roughly the same (homogeneity of variance)

March 14, 2005Methodology & Statistics This experiment o6 conditions testing 2 independent variables oDependent variable: reading time

March 14, 2005Methodology & Statistics o a self-paced reading study in Dutch o “who” and “which” questions o specificity and structure This experiment (2)

March 14, 2005Methodology & Statistics Why? oIn various research it is observed that “which” questions are more difficicult than “who” questions, but: oIt has never been addressed why this is the case

March 14, 2005Methodology & Statistics “Who”Wie heeft de keizer gezocht in de kelder? “Which” genericWelke persoon heeft de keizer gezocht in de kelder? “Which” specificWelke bediende heeft de keizer gezocht in de kelder? Variable 1: specificity

March 14, 2005Methodology & Statistics o Sentence structure is manipulated by context: SO Terwijl de dronken bediende een dutje deed, zocht de nuchtere bediende de keizer in de kelder. OS Terwijl de dronken bediende een dutje deed zocht de keizer de nuchtere bediende in de kelder. Variable 2: structure

March 14, 2005Methodology & Statistics o 3x2 design: 6 conditions WhoSOOS WhichGenSOOS WhichSpecSOOS o 8 scenario/question combinations per conditions (total of 48) Overview conditions

March 14, 2005Methodology & Statistics o 48 (14 male, 34 female) o Mean age 22.1 (sd 2.34) o Normal or corrected to normal vision o Paid for participation Participants

March 14, 2005Methodology & Statistics o Phrase-by-phrase self-paced reading (using E-prime software package) o“Moving window” o Accuracy: participants had to judge a provided answer (correct/incorrect) by pressing the corresponding button o Reading times and accuracy analyzed Procedure & analysis

March 14, 2005Methodology & Statistics Terwijl de dronken bediende een hapje at, zocht de nuchtere bediende de keizer in de kelder.

March 14, 2005Methodology & Statistics +

March 14, 2005Methodology & Statistics Wie

March 14, 2005Methodology & Statistics heeft

March 14, 2005Methodology & Statistics de keizer

March 14, 2005Methodology & Statistics gezocht

March 14, 2005Methodology & Statistics in de kelder?

March 14, 2005Methodology & Statistics

March 14, 2005Methodology & Statistics de nuchtere bediende

March 14, 2005Methodology & Statistics Hypotheses oSet-restriction (“specificity”) is a (the?) complicating factor during wh-question processing oAlso when the questions are presented within an appropriate context oBUT oProcessing difficulties in set-restricted wh- questions interact with difficulties in sentence structure

March 14, 2005Methodology & Statistics Hypothesis 1 oSet-restriction is a complicating factor during wh-question processing  WhichSpec > WhichGen = Who Alternatives:  WhichSpec = WhichGen > Who  WhichSpec = WhichGen = Who (in reading times (RTs)

March 14, 2005Methodology & Statistics Hypothesis 2 oProcessing difficulties in set-restricted wh- questions interact with difficulties in sentence structure  WhichSpec OS > WhichGen OS = Who OS  WhichSpec SO = WhichGen SO = Who SO Possible alternative:  All wh-types OS > SO

March 14, 2005Methodology & Statistics Segment of interest o“Point of integration”: Welke bediende heeft de keizer gezocht in de kelder? oAt this position (i.e. participle) it becomes clear which role each individual NP (Wh- phrase and “de keizer”) plays

March 14, 2005Methodology & Statistics Repeated measures/ SPSS

March 14, 2005Methodology & Statistics Data

March 14, 2005Methodology & Statistics Data pre-processing (1) oDefine conditions (in terms of factors) oCond 1 or 2: Qtype 1 (Who) oCond 1 or 4 or 5: Order 1 (SO) oDefine item groups oDefine lists

March 14, 2005Methodology & Statistics Data pre-processing (2) oDefine cut-offs (outliers) oCalculate segment means and sds oPer subject oPer item oDefine limits (mean+ 2sd) oReplace outliers and >limits by mean+2sd oData-transformation (suitable for SPSS)

March 14, 2005Methodology & Statistics Data pre-processing (3) oWrite syntax script

March 14, 2005Methodology & Statistics Data pre-processing (4) oWrite syntax script

March 14, 2005Methodology & Statistics Data pre-processing (5) oWrite syntax script

March 14, 2005Methodology & Statistics GLM – Repeated measures oSPSS data file with subject means oSPSS data file with item means

March 14, 2005Methodology & Statistics GLM – Repeated measures

March 14, 2005Methodology & Statistics GLM – Repeated measures

March 14, 2005Methodology & Statistics GLM – Repeated measures

March 14, 2005Methodology & Statistics GLM – Repeated measures

March 14, 2005Methodology & Statistics GLM – Repeated measures

March 14, 2005Methodology & Statistics Effects (subject analysis) oMain effect of question type (who, whichGen, whichSpec) oMain effect of order (SO, OS) oInteraction question type by order o(Interaction …by list)

March 14, 2005Methodology & Statistics Effect or interaction? oPlot you data! oMistake in labelling oEffect can be counter-intuitive oPost-hoc analyses

March 14, 2005Methodology & Statistics Accuracy

March 14, 2005Methodology & Statistics o The data suggest that answers following OS structures are more difficult to judge that those following SO questions. Accuracy

March 14, 2005Methodology & Statistics Reading times

March 14, 2005Methodology & Statistics o Phrase 1: wh-element Reading times (2)

March 14, 2005Methodology & Statistics o Phrase 2: auxiliary Reading times (3) ns **

March 14, 2005Methodology & Statistics o Phrase 3: NP2 Reading times (4) * ns *

March 14, 2005Methodology & Statistics o Phrase 4: Participle Reading times (5)

March 14, 2005Methodology & Statistics What do I want to know? oSo the separate question types behave differently? oIs this connected to word order complexity?

March 14, 2005Methodology & Statistics o Phrase 4: Participle Reading times (5)

March 14, 2005Methodology & Statistics o Phrase 4: Participle Reading times (5)

March 14, 2005Methodology & Statistics o Phrase 4: Participle Reading times (5)

March 14, 2005Methodology & Statistics What do I want to know? oSo the separate question types behave differently?  Who and WhichGen seem to pattern alike, compared to WhichSpec

March 14, 2005Methodology & Statistics Reading times SO

March 14, 2005Methodology & Statistics Reading times OS

March 14, 2005Methodology & Statistics What do I want to know? oIs a different pattern for the WhichSpec condition only connected to word order complexity?  The increased reading times for WhichSpec are confined to the OS structure-conditions. In SO conditions the average reading times were comparable.

March 14, 2005Methodology & Statistics o Phrase 5: PP Reading times (6)