Диссертация (1136198), страница 21
Текст из файла (страница 21)
As a result, we had 32participants, eight for each experimental list.After three participants were excluded, on average, 1.5% RTswere excluded as outliers (never more than 3.1% per region andcondition). Average RTs per region in different conditions arepresented in Figure 2.2 x 2 Repeated Measures ANOVAs with grammaticalityand gender match as factors were used to analyze RTs, as inExperiment 2a. Significant results were found only in regions5 (adjective/participle) and 6–7 (spillover regions).
They arepresented in Table 8.4.4.1. Masculine Head, Feminine AttractorThe main effect of Grammaticality was significant in analysis bysubjects and by items in regions 5–6, and only in analysis bysubjects in region 7. This reflects the fact that ungrammaticalsentences were read slower than grammatical ones. The maineffect of Gender Match was significant only in analysis by subjectsin regions 5–7. The interaction between Grammaticality andGender Match was not significant in any region.4. EXPERIMENT 2BIn this experiment we follow up on potential frequency effects inthe conditions with M heads from Experiment 2a.4.4.2.
Masculine Head, Neuter AttractorThe results were almost the same as in the other set of conditions.The main effect of Grammaticality was significant in regions5–7. The main effect of Gender Match was significant only inanalysis by subjects in regions 5–7. The interaction between thefactors never reached significance.4.1. ParticipantsThirty-five native Russian speakers (17 female, 18 male) took partin the experiment.
Ages ranged from 21 to 47 (mean age 31.3,SD 6.2).Frontiers in Psychology | www.frontiersin.org1250November 2016 | Volume 7 | Article 1651Slioussar and MalkoGender Agreement Attraction in RussianFIGURE 2 | Plots of mean RTs (in ms) by conditions in Experiment 2b. Error bars represent standard errors of the means. Regions: NP1 (1) - preposition (2) NP2 (3) - copula byt’ (4) - Adj/Part (5) - spillover (6–9). Ungrammatical conditions are red, grammatical ones are blue. Conditions where the gender of the attractor andthe predicate coincide (for example, MMM and MFF) have dark colors, conditions where they do not (for example, MFM and MMF) have light colors. (A) Masculinehead, feminine and masculine attractors, (B) Masculine head, neuter and masculine attractors.TABLE 8 | Results of the analysis for Experiment 2b.ConditionsRegionMM vs.
MF567MM vs. MN567FactorsdfMSeffectF1pdfMSeffectF2pGram1.31231038.5317.55<0.011.15104660.3415.92<0.01GenMatch1.3119014.9416.92<0.011.159218.40.930.35Gram * GenMatch1.311974.280.660.421.15887.30.130.72Gram1.3142715.9939.18<0.011.1523558.4110.520.01GenMatch1.3111614.799.320.011.156878.631.880.19Gram * GenMatch1.311445.201.070.311.15688.410.220.65Gram1.319604.9812.30<0.011.1547611.460.25GenMatch1.3111881.1120.57<0.011.157267.562.650.12Gram * GenMatch1.314900.502.890.101.153825.421.190.29Gram1.31137333.8533.02<0.011.1579129.6910.120.01GenMatch1.3114897.543.580.071.1510686.391.660.22Gram * GenMatch1.3110.64<0.010.961.15129.960.020.89Gram1.3175044.2214.43<0.011.1536864.008.800.01GenMatch1.313404.161.340.261.151161.110.370.56Gram * GenMatch1.312087.390.880.361.15704.90.120.73Gram1.3157868.0223.86<0.011.1532761.008.150.01GenMatch1.3110235.237.270.011.155016.181.250.28Gram * GenMatch1.311362.420.530.471.15419.230.10.76Analyses with p ≤ 0.05 are shown in bold.4.5.
Discussiongrammatical and ungrammatical conditions. In case of illusions,a different pattern would be expected: gender mismatch betweenthe head and the attractor should increase RTs in grammaticalconditions and decrease RTs in ungrammatical ones. Rather, itcan be suggested that gender mismatch carries some processingcost in the sentences with M heads. In any case, our data donot allow for strong claims: the main effect Gender Match issignificant in by subjects analysis in regions 5–7, but neverreaches significance in by items analysis.Since the outcome of comprehension experiments was notparallel to the results of Experiment 1 and earlier experimentson Slovak (Badecker and Kuminiak, 2007), we decided to lookThe results of this experiment show that the basic findingfrom Experiment 2a holds: there is no evidence for agreementattraction in the sentences with M heads.
The plots of the dataalso suggest that the unbalanced frequencies in Experiment 2ahad some influence on reading times. In Experiment 2b, wherethis confounding factor was excluded, two ungrammatical andtwo grammatical conditions pattern more closely together withineach condition set. Still, the conditions where the genders ofheads and attractors are mismatched have longer RTs.Notably, this difference in RTs is not an instance ofungrammaticality illusion, since it is observed in bothFrontiers in Psychology | www.frontiersin.org1351November 2016 | Volume 7 | Article 1651Slioussar and MalkoGender Agreement Attraction in Russianat the remaining combinations of head and attractor genders inExperiment 3 before suggesting an explanation.TABLE 9 | Frequencies of the attractors used in Experiment 3 (in ipm, orinstances per million).5.
EXPERIMENT 3In this experiment, we studied sentences with N heads and N, F,and M attractors and sentences with F heads and F, N, and Mattractors in comprehension. NF and FN combinations have notbeen examined before, and we added M attractors to be able tocompare sentences with all possible attractors.Head genderAttractor genderFF83.2M76.6N79.8NMean attractor freq (ipm)F92.1M92.8N95.75.1. ParticipantsIn Experiments 2a and 2b, we observed agreement attractionfor some combinations of head and attractor genders (FMand NM), but not for the others (MF and MN). So the firstquestion we asked in this experiment was whether there would beagreement attraction in NF and FN combinations.
If the answerwas yes, we were going to compare N and F attractors to Mattractors. To answer the first question, we took two groups ofconditions: FFF, FFN, FNF, FNN and NNN, NNF, NFN, NFF,and analyzed RTs using 2 x 2 Repeated Measures ANOVAs withgrammaticality and gender match as factors, as in the previousexperiments. Significant results were found only in regions 5(adjective or participle) and 6–7 (a spillover region). They arepresented in Table 10.Thirty-nine native Russian speakers (22 female, 17 male) tookpart in the experiment. Ages ranged from 19 to 40 (mean age 25.4,SD 6.4).5.2.
MaterialsWe constructed 36 sets of stimuli according to the same schemaas in Experiments 2a and 2b and observing the same constraints.Half of the sets had F head nouns and the other half hadN head nouns. In all sets, we used M, N, and F attractors.Their frequency was closely matched inside the three groupsof conditions, as Table 9 shows. Half of the predicates weregrammatical, and half were not. As a result, every target sentenceappeared in six conditions: NNN, NNF, NMN, NMM, NFN,NFF for the sentences with N heads and FFF, FFN, FMF,FMM, FNF, FNN for the sentences with F heads. Thus, outof all possible combinations of head, attractor and predicategenders, we did not use NNM and FFM.
We decided to doso to keep the number of grammatical and ungrammaticalconditions equal and sacrificed two conditions without anypotential for agreement attraction that we have already lookedat in Experiment 2a. Additionally, we used 100 fillers fromExperiment 2a. Experimental sentences were distributed intosix experimental lists, with factors counterbalanced. As aresult, we had 136 sentences per list (18 ungrammatical and118 grammatical), making the grammatical-to-ungrammaticalratio 6.6:1.5.4.1.
Feminine Head, Neuter AttractorThe main effect of Grammaticality was significant in analysisby subjects and by items in regions 5–6. This reflects the factthat ungrammatical sentences were read slower than grammaticalones. The main effect of Gender Match was significant onlyin analysis by subjects in regions 5–6. The interaction ofGrammaticality and Gender Match was significant in analysis bysubjects and by items in region 6 and only in analysis by subjectsin region 5. Ungrammatical sentences were read faster if the headand the attractor were mismatched in gender (i.e., in the FNNcondition compared to the FFN condition). This is the classicalattraction pattern, also known as a grammaticality illusion.
Atthe same time, there are no differences between grammaticalconditions, i.e., no evidence of ungrammaticality illusions wasfound.5.3. ProcedureThe procedure was the same as in Experiments 2a and 2b. Anexperimental session lasted around 11 min.5.4.2. Neuter Head, Feminine AttractorThe results were the same as in the other set of conditions.Thus, the answer to our first experimental question was positive,so we proceeded to compare the size of the attraction effectfor attractors of different genders. We compared two groupsof conditions: FNF, FNN, FMF, FMM and NFN, NFF, NMN,NMM.
We used 2 x 2 Repeated Measures ANOVAs withgrammaticality and attractor gender as factors. Only the maineffect of Grammaticality in region 6 was statistically significant[for conditions with F heads, F1 (1, 35) = 19.31, p < 0.01,MSeffect = 86064.00; F2 (1, 17) = 10.17, p = 0.01, MSeffect =24457.35; for conditions with N heads, F1 (1, 35) = 55.80,p < 0.01, MSeffect = 126973.44; F2 (1, 17) = 7.32, p = 0.02,MSeffect = 52915.47]. The main effect of Attractor Gender or theinteraction between the factors were not significant in any region.5.4.
ResultsWe analyzed participants’ question-answering accuracy andreading times. The data from three participants were discardedbecause they had <75% accuracy in comprehension questions.As a result, we had 36 participants, six for each experimental list.None of them made more than two mistakes when answeringquestions to target sentences (i.e., 12.5% at most).As in the previous experiments, reading times that exceededa threshold of 2.5 standard deviations, by region and condition,were excluded.
In total, 1.8% of the data were excluded (nevermore than 3.7% per region and condition). Average RTs perregion in different conditions are presented in Figure 3 (noticethat coloring conventions are different from the previous plots).Frontiers in Psychology | www.frontiersin.org1452November 2016 | Volume 7 | Article 1651Slioussar and MalkoGender Agreement Attraction in RussianFIGURE 3 | Plots of mean RTs (in ms) by conditions in Experiment 3. Error bars represent standard errors of the means. Regions: NP1 (1) - preposition (2) - NP2(3) - copula byt’ (4) - Adj/Part (5) - spillover (6–9).
![](https://s.studizba.com/z.php?f=/templates/si/i/noavatar.png&w=100&h=100&t=1"){kind=avatar}
