Диссертация (1136198), страница 20
Текст из файла (страница 20)
For twoparticipants, this led to the exclusion of more than 15% responses,so we did not include their data in further analysis.After four participants were excluded, we had 44 participants(11 in each experimental list). In total, 2.3% of the data wereexcluded as outliers (never more than 3.6% per region andcondition). Average RTs per region in different conditions arepresented in Figure 1.The data for each set of conditions (e.g., MMM - MFM MMF - MFF) were entered in a 2 × 2 Repeated MeasuresANOVA with grammaticality and gender match between theattractor and the head nouns as factors.
We used IBM SPSSsoftware (www.ibm.com/software/analytics/spss/). Analyses byitems and by participants were performed. Data from all regionswere tested, but there were significant results only in regions4–6 in the conditions with M heads and in regions 5–6 in theconditions with F and N heads. Region 4 is the copula, region 5is an adjective or participle, regions 6–10 contain several wordsFrontiers in Psychology | www.frontiersin.org3.4.1. Feminine Head, Masculine AttractorThe main effect of Grammaticality is significant in analysis bysubjects and by items in regions 5–6, reflecting the fact thatungrammatical sentences were read slower than grammaticalones.
The main effect of Gender Match is not significant in anyregion. The interaction of Grammaticality and Gender Match issignificant in analysis by subjects and by items in region 5 andonly in analysis by subjects in region 6. Ungrammatical sentenceswere read faster if the head and the attractor were mismatchedin gender (i.e., in the FMM condition compared to the FFMcondition).
This is the classical attraction pattern.3.4.2. Neuter Head, Masculine AttractorThe main effect of Grammaticality is significant in regions 5–6,reflecting longer RTs in ungrammatical conditions. The maineffect of Gender Match is significant only in analysis by subjectsin regions 5–6. The interaction of Grammaticality and GenderMatch is significant in regions 5–6, which is again a reflectionof the classical attraction pattern: NMM condition was readfaster than NNM and, in fact, almost as fast as grammaticalconditions.1048November 2016 | Volume 7 | Article 1651Slioussar and MalkoGender Agreement Attraction in RussianTABLE 6 | Results of the analysis for Experiment 2a.ConditionsFF vs. FMRegion56NN vs.
NM56MM vs. MF456MM vs. MN456FactordfMSeffectF1pdfMSeffectF2pGram1.43103775.6418.29GenMatch1.435355.951.56<0.011.1135056.8317.86<0.010.221.113326.672.05Gram * GenMatch1.4350162.640.1820.54<0.011.1121717.525.010.05<0.01Gram1.4362551.8421.23<0.011.1116335.6318.42GenMatch1.430.05<0.011.001.110.05<0.011.00Gram * GenMatch1.4314823.134.650.041.113398.652.760.13Gram1.4378213.5528.70<0.011.1120398.138.870.01GenMatch1.4333363.0610.93<0.011.119996.531.920.19Gram * GenMatch1.4336720.3516.67<0.011.1111405.2529.57<0.01<0.01Gram1.4371017.3532.99<0.011.1118794.1727.45GenMatch1.4351758.7026.98<0.011.1112558.271.990.19Gram * GenMatch1.4320026.3114.92<0.011.114945.087.400.02Gram1.438423.585.210.031.112100.132.120.17GenMatch1.4393656.8258.58<0.011.1125002.518.460.01Gram * GenMatch1.4370.010.050.831.110.05<0.010.99Gram1.4363205.8316.87<0.011.1119008.4816.83< 0.01GenMatch1.4334672.6310.56<0.011.119163.211.110.32Gram * GenMatch1.43114.410.020.881.1119.250.020.89Gram1.4332730.0013.37<0.011.118554.689.060.01GenMatch1.4347491.2527.90< 0.011.1112185.812.240.16Gram * GenMatch1.432128.790.880.351.11401.361.170.30Gram1.431264.210.750.391.11231.000.380.55GenMatch1.4366406.3124.34<0.011.1115699.952.970.11Gram * GenMatch1.432116.291.880.181.11321.891.340.27Gram1.4363247.5320.26<0.011.1116965.1223.11<0.01GenMatch1.4386314.1218.61<0.011.1122733.112.230.16Gram * GenMatch1.43414.210.130.721.111.02< 0.010.96Gram1.4336279.689.04<0.011.119509.077.940.02GenMatch1.4352540.1914.24<0.011.1115123.001.760.21Gram * GenMatch1.4329.050.010.931.1140.330.020.90Analyses with p ≤ 0.05 are shown in bold.Gender Match is significant only in analysis by subjects in regions4–6.
The interaction of Grammaticality and Gender Match isnot significant in any region, so these conditions also show noagreement attraction.3.4.3. Masculine Head, Feminine AttractorThe main effect of Grammaticality is significant in analysis bysubjects in region 4 and in analysis by subjects and by itemsin regions 5–6. This reflects the fact that RTs were longer inungrammatical conditions. The main effect of Gender Match issignificant in analysis by subjects and by items in region 4, andonly in analysis by subjects in regions 5–6.
This correspondsto longer RTs in conditions where the genders on the nounswere mismatched. The interaction of Grammaticality and GenderMatch did not reach significance in any regions, which points tothe absence of agreement attraction.3.5. DiscussionAs can be seen from the analyses, the results fall into two groups.In the conditions with F or N heads and M attractors thereis clear evidence for gender agreement attraction.
RTs exhibitthe classical attraction profile with grammaticality illusions:ungrammatical sentences where the attractor and the predicatehave the same gender (FMM and NMM) are read faster thanother ungrammatical sentences (FFM and NNM). Discussingcomprehension studies of number agreement attraction inthe introduction, we outlined different approaches to thisphenomenon, but will opt for one of them ourselves only in3.4.4.
Masculine Head, Neuter AttractorThe main effect of Grammaticality is significant in analysisby subject and by items in regions 5–6: the ungrammaticalconditions are read slower than grammatical. The main effect ofFrontiers in Psychology | www.frontiersin.org1149November 2016 | Volume 7 | Article 1651Slioussar and MalkoGender Agreement Attraction in Russian4.2.
MaterialsTABLE 7 | Frequencies of the attractors used in Experiments 2a and 2b (inipm, or instances per million).ExperimentExperiment 2aHeadAttractorgendergenderFF138.1M120.0N105.9M81.7NMMExperiment 2bMMWe constructed 32 sets of stimuli according to the same schemaas in Experiment 2a and observing the same constraints. Headnouns were always masculine. In 16 sets, the attractors weremasculine and neuter; in the other 16 sets, the attractors weremasculine and feminine.
Most of the head nouns were re-usedfrom the Experiment 2a, but we replaced attractors so thattheir frequencies were closely matched inside the two groupsof conditions. We used The Frequency Dictionary of ModernRussian Language (Lyashevskaya and Sharoff, 2009). Averagefrequencies of head and attractor nouns in Experiments 2a and2b are shown in Table 7. As in Experiment 2a, half of thepredicates did not agree with the subject in gender. Additionally,we used 80 fillers from Experiment 2a. Experimental sentenceswere distributed into four experimental lists, with factorscounterbalanced. As a result, we had 112 sentences per list (16ungrammatical and 96 grammatical), making the grammaticalto-ungrammatical ratio 6:1.Mean attractorfreq (ipm)M91.8F41.1M134.9aN78.9M61.4F61.9M69.2N68.14.3.
ProcedureThe procedure was the same as in Experiment 2a. Anexperimental session lasted around 9 min.a Itshould be noted that one really frequent M noun influences this number a lot. If we getrid of it and of the corresponding N attractor, the frequencies become very close: 73.9 forM attractors and 84.6 for N attractors.4.4. Resultsthe general discussion section once all experimental findingsare presented. Let us also note that ungrammaticality illusionsare absent: in the sentences with N heads there are virtuallyno differences between grammatical conditions; in the sentenceswith F heads, they are insignificantly small.On the other hand, the conditions with M heads and For N attractors do not show any evidence of attraction.
Bothgrammatical and ungrammatical sentences where the head andthe attractor match in features (MMM, MMF, and MMN) areread faster than the sentences where they are mismatched (MFM,MNM, MFF, and MNN). In case of ungrammatical sentences, thispattern is the reverse of what we usually see in attraction cases.Looking for an explanation of such pattern, we discoveredthat we need to rule out an important confound first.Unfortunately, we made a mistake during the preparationof experimental materials, and the frequencies of attractorsin conditions with M heads were not well balanced. Sincethis could influence the results in some unexpected way, weconducted an additional experiment where the frequencieswere carefully controlled.
Conditions with F and N headsdid not have this problem, and the results reported forthem hold.Like in Experiment 2a, we analyzed participants’ questionanswering accuracy and reading times. At the first stages ofanalysis, the data from three participants were discarded: one ofthem had <75% accuracy in comprehension questions; the othertwo read too slowly compared with the others, so more than 15%of their RTs would have to be excluded as outliers (exceedingthe threshold of 2.5 standard deviations).
![](https://s.studizba.com/z.php?f=/templates/si/i/noavatar.png&w=100&h=100&t=1"){kind=avatar}
