CC BY
29
Katharina Korecky-Kroll
Department of Linguistics, University of Vienna
ENCOURAGEMENTS AND DISCOURAGEMENTS
IN PARENTAL INPUT: EVIDENCE FROM HIGH AND LOW SES FAMILIES
1. Social differences in parenting styles
This paper aims at discovering overall tendencies of conversation styles in German-speaking families from different socioeconomic backgrounds. The focus will be on encouragements and discouragements, as studies on English-speaking populations have shown that this is an area in which language input to young children may show considerable differences due to the socioeconomic status (SES) of families.
Many studies on language acquisition (Ensminger & Fothergill 2003, Hoff 2006) assess SES of children via the highest level of parental (most often: maternal) education: High SES (HSES) parents mostly hold college or university degrees, low SES (LSES) parents have often only compulsory schooling. Some studies also include family income (Rowe 2008), prestige of parental profession as opposed to unemployment (Hart & Risley 1995) or the total family capital (consisting of financial, human, and social capital, cf. Chiu and McBride-Chang 2006).
SES has been shown to be a decisive factor for children's linguistic and cognitive development (Hart & Risley 1995, Brito and Noble 2014) and also for their later school performance (Walker et al. 1994): Children from LSES families have smaller vocabularies (HoffGinsberg 1998), show a slower phonological, morphological and syntactic development (Bowey 1995, Ravid 1995, Huttenlocher et al. 2010) and poorer performance not only in linguistic tests, but also in neurocognitive processing studies (Noble et al. 2005).
Parental language input is considered to be the most important mediating variable between SES and children's linguistic proficiencies (e.g. Huttenlocher et al. 2010): In their seminal study on the language
experience of 42 American children aged 1-3, Hart and Risley (1995) demonstrate that by age 3, children from professional (= HSES) families have had 30 million words of cumulative experience more than children from welfare (= the lowest SES) families, whereas the working-class children (LSES, but not as low as the welfare children) were in between. The children's cumulative language experience was clearly reflected in their vocabularies and IQ scores at age 3.
In addition, HSES parents' child-directed speech differs not only in quantity, but also in quality from that of LSES parents (Hoff 2003): According to various studies on mostly English-speaking populations (Hart & Risley 1995, Hoff-Ginsberg 1991, 1998, Hoff et al. 2002), parental conversation style is closely related to SES of families: Parents from HSES backgrounds, who mostly have broader knowledge about child development and child care issues (Rowe 2008), are more responsive to their children's verbalizations, initiate and sustain conversation with their children more frequently and encourage them more often to talk by asking them questions (Hoff 2003). HSES parents also tend to formulate their requests in an indirect way, e.g., as questions, such as "Why don't you pick up the toys for me?" (cf. Hart & Risley 1995: 57).
Otherwise, parents from LSES backgrounds, who often experience greater social stress and are thus more focused on goal-directed caretaking settings than on playing situations, use more behavior-directing speech acts (Hoff-Ginsberg 1991), such as direct commands and prohibitions ("Put it here!", "Don't touch it!").
Although commands and prohibitions may be regarded as particularly typical for LSES parents' speech, they are not the only speech acts that may be responsible for an overall atmosphere of discouragement that is often found in LSES families. For example, Hart and Risley (1995: 199, 253) classify affirmations that repeat or expand the child's utterances as well as expressions of approval such as "right" as encouragements, whereas they consider prohibitions, commands like "shut up", but also expressions of disapproval such as "bad" as discouragements. On the basis of their impressive longitudinal data of 42 American children from three socioeconomic backgrounds (professional, working-class, welfare), Hart & Risley (1995: 198-199) demonstrate that an average child in a professional family hears 32 affirmatives and 5 prohibitions per hour, whereas a child in
a working-class family hears 12 affirmatives and 7 prohibitions and a child growing up in a welfare family hears only 5 affirmatives, but 11 prohibitions per hour. If this is extrapolated to the cumulative language experience of the child's first four years of life, a child in a professional family will have heard 560,000 more instances of encouragements than of discouragements, whereas a child in a welfare family will have heard 125,000 more discouragements than encouragements (Hart & Risley 1999: 199), which is clearly a massive difference in positive vs. negative feedback.
2. Encouraging and discouraging response particles and expressive speech acts
As it is not possible to investigate all types of encouragements and discouragements in the present study1, the analysis will be limited to encouraging vs. discouraging response particles as well as to encouraging vs. discouraging expressive speech acts.
The response particles2 yes and no (German ja and nein) and their less formal equivalents uh-huh (German: mhm, meaning yes) and uh-uh (German: mm, meaning no) represent one simple dichotomy of encouragement vs. discouragement. One analysis of this paper will thus consist in contrasting token frequencies of ja and nein and of their informal equivalents in families of two different socio-economic backgrounds.
Expressive speech acts (cf. Searle & Vanderveken 1985: 58) "express the speaker's attitude" or feelings about a state of affairs3.
1 A chapter on the use of requests in the same families has already been submitted to the editors of a volume on the acquisition of Modality (cf. Korecky-Kroll, submitted). Although requests can often be regarded as discouraging directive speech acts, they will be investigated in the chapter of the Modality volume and disregarded in the present paper.
2 Although there is some disagreement about the exact part-of-speech classification of yes and no (they are sometimes classified as interjections, sometimes as adverbs, sometimes as sentence words), we will adopt the terminology of Watts (1986: 166), who classifies them as response particles, a notion whose German equivalent "Antwortpartikel" is also used in the Duden Grammatik (2006: 603).
3 Other speech acts are that are found in child speech and child-directed speech are assertives (e.g. assertions and statements), directives (e.g. requests
Typical expressive speech acts are praises, but also disapprovals, complaints, greetings, curses, apologies, expressions of surprise etc. Not every expressive speech act is of course encouraging or discouraging, but praises can be regarded as clear encouragements (at least if they are not used with a sarcastic undertone), whereas disapprovals, complaints and curses can be considered to be discouraging. In this chapter, we will focus on encouraging and discouraging expressive speech, whereas we will not investigate "neutral" expressive speech acts, speech acts that are neither encouraging nor discouraging (e.g. a mere expression of surprise like oh!). Examples (1) and (2) represent typical encouraging speech acts (the first is from a mother, the second from a child), whereas examples (3) and (4) are likely to have a discouraging perlocutionary effect on the addressee. Please note that examples (1) and (2) are from the same parent-child dyad and examples (3) and (4) are from another parent-child dyad:
(1) *PAR: du hast eh den Tisch
you.2SG have.PRS.2SG anyway the.ACC.MASC table schon so schon dekoriert !
already so beautiful decorate.PAST.PTCP ! 'you have already decorated the table so beautifully anyway!' (HSES mother to her daughter, aged 4;6)
(2) *CHI: du kannst urgut tanzen, Mama!
you.2SG can.PRS.2SG very.well dance.INF Mom ! 'you can dance very well, Mom!' (HSES girl, aged 3;7, to her mother)
(3) *PAR: Sucht-ler !
addict-PEJ.MASC ! 'junkie!'
(LSES mother to her son, aged 3;2, who is watching TV)
and questions), commissives (e.g. promises, offers and threats). Searle and Vanderveken (1985: 37ff.) distinguish them according to the way of interference with the addressee's world of actions: While assertives say how things are (direction of fit: word to world), both directive and commissive speech acts want to change the world (direction of fit: world to word) by getting either the hearer (in directives) or the speaker (in commissives) to perform an action. Expressive speech acts show an empty direction of fit.
(4) *CHI: du bist eine Kuh schon !
you.2SG be.PRS.2SG a.NOM.FEM cow oh yes !
'oh yes, you are a cow!'
(LSES boy, aged 4;5, to his mother)
In accordance with previous studies on differences in parenting styles (see section 1), we hypothesize that HSES parents will use more encouraging response particles as well as more encouraging expressive speech acts in comparison to LSES parents. As children tend to follow their language input, we assume that HSES children will also use more encouraging response particles and more encouraging expressive speech acts in comparison to LSES children. Nevertheless, as all children's needs are similar — irrespective of their SES — these differences may be smaller in the children than in the parents.
3. Method
3.1. Participants
29 German-speaking parent-child dyads living in Vienna (Austria) were recorded at home at four data points: The children had a mean age of 3;1 (age range: 2;11-3;3) at the beginning of the study and mean ages of 3;4, 4;4 and 4;8 at the follow-up recordings (see also 3.2). The groups were nearly balanced for SES and gender (see Table 1).
Table 1. Child participants
Child SES Child gender N of children Subtotal SES
HSES female 8 15 HSES
HSES male 7
LSES female 6 14 LSES
LSES male 8
Total 29
Following other studies on language acquisition (cf. Ensminger & Fothergill 2003), SES was mainly assessed by the main parental caretaker's highest educational level (cf. OECD 1999): The LSES group included ISCED-97 levels 1 to 3b (i.e. from compulsory school to apprenticeship and vocational schools, but without high school diploma), whereas the HSES group had ISCED-97 levels 3a to 6 (i.e. from high school diploma up to PhD).
The prestige of the parental profession was assessed according to the International Socioeconomic Index of Occupational Status (ISEI, cf. Ganzeboom & Treiman 1996). Nevertheless, the assessment of the ISEI values affected the SES classification of only one child in the sample: One boy's mother had a significantly better job than would have been expected from her formal educational level. Therefore, this boy was "upgraded" to the HSES group4.
In contrast to Hart & Risley (1995), who investigated three different SES groups (namely professional, working-class and welfare, see also section 1), the LSES families of our study were all working-class families — we did not investigate the lowest SES group of welfare families because we were unable to find a sufficient number of participants.
3.2. Procedure
We conducted four one-hour spontaneous speech audio and video recordings in each child's home (Rowe 2012) within a period of approximately 1 '/2 years (at children's mean ages 3;1, 3;4; 4;4 and 4;8, see section 3.1). Parents were not given any detailed instructions, but they were asked to do what they usually did. Therefore, situations were as natural as they could be (despite the observer's paradox, which was the same for everyone, cf. Labov 1972), but they also showed considerable variation: Some parents asked their children to choose a game, others read storybooks, others just had spontaneous conversation. In a few cases, we also recorded mealtime situations (cf. Hoff-Ginsberg 1991). Sometimes, siblings or other adults (e.g. the other parent, a grandparent, an aunt or a visitor) were also present, whereas in other cases, the conversations were limited to the parent-child dyads.
From each of these one-hour recordings, we selected the 30 minutes with the richest verbal parent-child interaction. The selected samples frequently consisted of two parts of different lengths that were summed up to exactly 30 minutes in total.
3.3. Transcription and coding
All 116 30-minute samples were transcribed according to the CHAT conventions of the CLAN program package (MacWhinney
4 For details on the SES classification of the children investigated in the present study cf. Czinglar et al. (2015).
2000) and tagged for parts-of speech and morphology by using the lexicon-based approach (MacWhinney 2000). According to this approach, new lexical entries and morphological forms were first identified, then manually coded and added to a lexicon file. The MOR program of the CLAN program package was then used to automatically generate a morphological coding tier after each transcription tier on the basis of the coding in the lexicon file. Afterwards, ambiguous word forms were manually disambiguated (Korecky-Kroll 2017). Finally, these morphological coding tiers were used to calculate token frequencies of the response particles described in section 2).
For the present study, we only considered spontaneous child speech and child-directed speech as well as citations (e.g., songs, rhymes or book reading), but children's imitations as well parents' utterances directed to other adults or to pets were excluded from the counting.
In order to insert the speech act codings (see categories described in section 2), we imported the morphologically coded files into MS Excel (Korecky-Kroll in press) by using the CLANTOCSV JavaScript program (Korecky 2015). Speech acts were first identified and coded according to their context in a separate column, and as a second step, another column was inserted to add a more detailed coding of encouraging vs. discouraging expressive speech acts. Frequencies of all relevant speech act categories were obtained via pivot tables.
3.4. Statistical analysis
We used the MASS package (Venables & Ripley 2002) of R (R Core Team 2015) to perform chi square tests of independence of the relationship between SES and encouraging vs. discouraging response particles as well as between SES and encouraging vs. discouraging expressive speech acts in order to discover group differences between parents and children. A post-hoc residual analysis was conducted to evaluate differences between specific cells (cf. Sharpe 2015).
4. Results
4.1. Encouraging and discouraging response particles
Chi-square tests of independence were performed to examine the relation between SES and encouraging vs. discouraging response particles, for parents and children separately (see also raw token frequencies and percentages in Tables 2 and 3). While the relation be-
tween these variables was significant for the parents (%2 = 78.394, df = 1, p < 0.001), this was not the case for the children (%2 = 0.006, df = 1, p = 0.937). The post-hoc analysis for the parents revealed that HSES parents used significantly more encouraging response particles and significantly fewer discouraging response particles, whereas LSES parents used significantly fewer encouraging response particles, but significantly more discouraging response particles.
Table 2. SES and encouraging vs. discouraging response particles (RPs)
in parents
encouraging RPs discouraging RPs Total
HSES LSES 2620 (72.42 %) 1598 (61.75 %) 998 (27.58 %) 990 (38.25 %) 3618 (100.00%) 2588 (100.00%)
Total 4218 (67.97 %) 1988 (32.03%) 6206 (100.00%)
Table 3. SES and encouraging vs. discouraging response particles (RPs) in children
encouraging RPs discouraging RPs Total
HSES LSES 1877 (56.43%) 1377 (56.57%) 1449 (43.57 %) 1057 (43.43 %) 3326 (100.00%) 2434 (100.00%)
Total 3254 (56.49%) 2506 (43.51 %) 5760 (100.00%)
4.2. Encouraging and discouraging expressive speech acts The following analysis uses the same methodology as discussed in section 4.1, but it examines the relationship between SES and encouraging vs. discouraging speech acts. It is thus an analysis based on frequencies of speech acts and not an analysis based on word token frequencies (in contrast to the analysis shown in 4.1).
The chi-square tests yielded significant results for both parents (X2 = 103.42, df = 1, p < 0.001) and children (x2 = 10.465, df = 1, p = 0.001), indicating a relationship between SES and discouraging vs. encouraging speech acts (see also raw frequencies and percentages of speech acts in Tables 4 and 5). Again, the post-hoc analysis for the parents revealed that HSES parents used significantly more encouraging speech acts and significantly fewer discouraging speech acts, whereas LSES parents used significantly more discouraging speech acts and significantly fewer encouraging speech acts. The same was
true for the children, but the differences were larger for the parents than for the children. Nevertheless, LSES children's expressive speech acts were the only category in which discouragements (53.17%) even prevailed over encouragements (46.83%).
Table 4. SES and encouraging vs. discouraging expressive speech acts (SAs)
in parents
encouraging SAs discouraging SAs Total
HSES LSES 1478 (67.64 %) 830 (51.30%) 707 (32.36 %) 788 (48.70 %) 2185 (100.00 %) 1618 (100.00 %)
Total 2308 (60.69 %) 1495 (39.31%) 3803 (100.00 %)
Table 5. SES and encouraging vs. discouraging expressive speech acts (SAs) in children
encouraging SAs discouraging SAs Total
HSES LSES 822 (52.93%) 628 (46.83%) 731 (47.07%) 713 (53.17%) 1553 (100.00%) 1341 (100.00%)
Total 1450 (50.10 %) 1444 (49.90%) 2894 (100.00%)
5. Discussion
The aim of this paper was to investigate large overall tendencies of conversation styles in German-speaking families of different socioeconomic backgrounds. The focus was on encouragements and discouragements, more precisely on encouraging vs. discouraging response particles (such as yes and no) as well as on encouraging vs. discouraging expressive speech acts.
In accordance with previous studies (e.g. Hart & Risley 1995), it was confirmed that HSES parents use a clearly more encouraging parenting style than LSES parents, which was reflected in a higher use of encouraging response particles and encouraging speech acts. Although encouraging response particles and encouraging expressive speech acts were more frequent than their discouraging counterparts in all parents, a tendency which is also shown by Hart & Risley (1995) with respect to professional and working-class parents 5. Of course, the
5 Remember that the opposite tendency was found for the lowest SES group of welfare parents investigated by Hart & Risley (1995). Average wel-
categories investigated were slightly different in both studies: Whereas Hart & Risley (1995) included prohibitions and commands, they were excluded from the present study6. Nevertheless, the German-speaking parents of the present paper showed a smaller SES gap than the English-speaking parents investigated by Hart & Risley (1995): Whereas the German-speaking HSES parents used only 67.64% encouraging vs. 32.36% discouraging expressive speech acts (as opposed to 86.49 % encouragements vs. 13.51% discouragements in the English-speaking HSES parents), the German-speaking LSES parents used 51.30 % encouraging vs. 48.70 % discouraging expressive speech acts (in contrast to 63.16% encouragements vs. 36.84 % in the working-class parents of the Hart & Risley study). Nevertheless, it is not possible to say whether these differences between the two studies reflect different cultural habits (i.e. a generally more encouraging North American child-rearing perspective as opposed to a more critical European perspective), whether they are due to the differences in the linguistic categories investigated or to the children's ages7.
As far as the children are concerned, we found no significant group differences with respect to encouraging vs. discouraging response particles: Thus, young children use very similar rates of yes and no as well as of their colloquial equivalents, irrespective of their SES, which reflects that both groups of children have basically the same needs.
Nevertheless, the speech act analysis, which was probably the more detailed of both analyses8, showed significant SES differences,
fare parents produced 5 encouragements and 11 discouragements per hour, which corresponds to a rate of 31.25% encouragements as opposed to 68.75% discouragements. Nevertheless, as we could not find a sufficient number of German-speaking welfare families, we do not have a comparison group in the present study.
6 But they will of course be investigated in a separate chapter (see Ko-recky-Kroll submitted).
7 The children in the Hart & Risley study were aged 1-3, whereas the German-speaking children were between ages 2;11 and 4;11. It is likely that parents behave in a more encouraging way when talking to younger children, whereas they are more critical when talking to older children.
8 Note that in the response particle analysis, we counted just frequencies of response particles, irrespective of their context (e.g. if they were really used in an encouraging way), whereas in the speech act analysis, we tried to evaluate each speech act in its pragmatic context.
which reflect the differences in the children's input. Clearly, a child who hears more encouragements, will find it more natural to use them as well, and the same holds of course for discouragements.
Overall, the SES differences were smaller in the children than in the parents, suggesting that SES-determined speech is a usage-based norm that children will grow into when getting older.
6. Conclusion
Although this study focused only on one small aspect of child-directed speech, namely on encouraging vs. discouraging response particles and expressive speech acts, it shows nevertheless important SES differences. Unfortunately, it seems to be the case that education is still inherited and that SES disparities continue to persist, despite all political measures that have been taken in order to attenuate SES effects of families on children's education. But apparently these measures are not yet sufficiently widespread or effective, as we would need much more educational and financial support for parents in the families as well as special education programs for disadvantaged children in kindergarten settings and of course, early obligatory kindergarten attendance for all children, not only for those whose parents have well-paid fulltime jobs.
7. Acknowledgements
First of all, I would like to thank all relevant institutions: Collection and transcription of the data of this study were financed by the Wiener Wissenschafts-, Forschungs- and Technologiefonds (WWTF, SSH11-027), whereas the positions of the author were financed by the University of Vienna (2012-2015) and the Austrian Academy of Sciences (2016).
Furthermore, I am deeply grateful to Wolfgang U. Dressler, the principal investigator of the INPUT project, for his continuous support, fruitful discussions and his permission to use the project data for this chapter. I also thank my colleagues Christine Czinglar, Sabine Sommer-Lolei, Viktoria Templ, Kumru Uzunkaya-Sharma, and Maria Weichselbaum for their great cooperation in various steps of data collection, preparation and analysis as well as numerous undergraduate students at the Department of Linguistics of the University of Vienna for their transcription work.
Among my international cooperation partners, I would like to express my thanks to Maria Voeikova and Kira Ivanova for inviting me to the lovely city of St. Petersburg to present my results at the conference "Modality in the Language of Children and Adults" and to the editors of the Modality volume, Ursula Stephany, Ayhan Aksu-Ko?, and Maria Voeikova for interesting discussions and helpful comments. I also thank Dorit Ravid who inspired me to start my research on low SES children and Basilio Calderone who gave me invaluable statistical advice.
Last, but not least, I thank all children and parents for their participation.
References
Bowey 1995 — J. Bowey. Socioeconomic status differences in preschool phonological sensitivity and first-grade reading achievement // Journal of Educational Psychology 87, 3, 1995. P. 476-487. Brito, Noble 2014 — N. H. Brito, K. G. Noble. Socioeconomic status and
structural brain development. Frontiers in Neuroscience 8, 276, 2014. Chiu, McBride-Chang 2006 — M. M. Chiu, C. McBride-Chang. Gender, context, and reading: a comparison of students in 43 countries // Scientific Studies of Reading 19, 2006. P. 331-362. Czinglar et. al. 2015 — C. Czinglar, K. Korecky-Kröll, K. Uzunkaya-Sharma, W. U. Dressler. Wie beeinflusst der sozioökonomische Status den Erwerb der Erst- und Zweitsprache? Wortschatzerwerb und Geschwindigkeit im NP/DP-Erwerb bei Kindergartenkindern im türkischdeutschen Kontrast // K.-M. Köpcke, A. Ziegler (eds.). Deutsche Grammatik in Kontakt. Deutsch als Zweitsprache in Schule und Unterricht. Berlin: Mouton de Gruyter, 2015. P. 207-240. Duden 2006 — Die Grammatik. Mannheim: Dudenverlag, 2006. Ensminger, Fothergill 2003 — M. E. Ensminger, K. E. Fothergill. A decade of measuring SES: What it tells us and where to go from here // M. H. Bornstein, R. H. Bradley (eds.). Socioeconomic status, parenting and child development. Mahwah, NJ: Erlbaum, 2003. P. 13-27. Ganzeboom, Treiman 1996 — H. B. G. Ganzeboom, D. J. Treiman. Internationally Comparable Measures of Occupational Status for the 1988 International Standard Classification of Occupations // Social Science Research 25, 1996. P. 201-239. Hart, Risley 1995 — B. Hart, T. Risley. Meaningful differences in the everyday
experience of young American children. Baltimore: Brookes, 1995. Hart, Risley 1999 — B. Hart, T. Risley. The Social World of Children: Learning to Talk. Baltimore, MD, Paul H. Brookes Publishing Co., 1999.
Hoff 2003 — E. Hoff. Causes and consequences of SES-related differences in parent-to-child speech // M. H. Bornstein, R. H. Bradley (eds.). Socioeconomic status, parenting, and child development. Mahwah, NJ: Erlbaum. P. 147-160.
Hoff 2006 — E. Hoff. How social contexts support and shape language development // Developmental Review 26, 2006. P. 55-88.
Hoff, Laursen, Tardif 2002 — E. Hoff, B. Laursen, T. Tardif. Socio-economic status and parenting // M. H. Bornstein (ed.), Handbook of Parenting. 2nd edn. Mahwah (NJ): Erlbaum, 2002. P. 231-252.
Hoff-Ginsberg 1991 — E. Hoff-Ginsberg. Mother-child conversation in different social classes and communicative settings // Child Development 62, 4, 1991. P. 782-796.
Hoff-Ginsberg 1998 — E. Hoff-Ginsberg. The relation of birth order and socioeconomic status to children's language experience and language development // Applied Psycholinguistics 19, 4, 1998. P. 603-629.
Huttenlocher et al. 2010 — J. Huttenlocher, H. Waterfall, M. Vasilyeva, J. Vevea, L. V. Hedges. Sources of variability in children's language growth // Cognitive Psychology 61, 2010. P. 343-365.
Korecky-Kröll 2017 — K. Korecky-Kröll. Kodierung und Analyse mit CHILDES: Erfahrungen mit kindersprachlichen Spontansprachkorpora und erste Arbeiten zu einem rein erwachsenensprachlichen Spontansprachkorpus // C. Resch, W. U. Dressler (eds.). Digitale Methoden der Korpusforschung in Österreich. Wien: Verlag der Österreichischen Akademie der Wissenschaften, 2017. P. 85-113.
Korecky-Kröll submitted — K. Korecky-Kröll. Requests in first language acquisition of German: Evidence from high and low SES families // U. Stephany, A. Aksu-Kog, M. D. Voeikova (eds.). Development of Modality in First Language Acquisition. Boston: De Gruyter, series: Studies on Language Acquisition. Submitted.
Korecky 2015 — Korecky, Paul C. 2015. CLANTOCSV [Computer software].
Labov 1972 — W. Labov. Sociolinguistic Patterns. Oxford: Blackwell, 1972.
MacWhinney 2000 — B. MacWhinney. The CHILDES Project: Tools for Analyzing Talk. 3rd edn. Mahwah, NJ: Erlbaum, 2000.
Noble, Norman, Farah 2005 — K. G. Noble, M. F. Norman, M. J. Farah. Neurocognitive correlates of socioeconomic status in kindergarten children // Developmental Science 8, 1, 2005. P. 74-87.
OECD 1999 — OECD. Classifying Educational Programmes. Manual for ISCED-97 Implementation in OECD Countries. OECD: 1999 Edition, 1999.
R Core Team 2015 — R Core Team. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria, 1999. URL http: www.R-project.org/.
Ravid 1995 — D. Ravid. Language change in child and adult Hebrew: a psy-cholinguistic perspective. New York: Oxford University Press, 1995.
Rowe 2008 — M. Rowe. Child-directed speech: relation to socioeconomic status, knowledge of child development and child vocabulary skill. Journal of Child Language 35, 2008. P. 185-205.
Rowe 2012 — M. Rowe. Recording, transcribing, and coding interaction // E. Hoff (ed.). Research methods in child language: a practical guide. Chichester: Wiley-Blackwell, 2012. P. 193-207.
Searle, Vanderveken 1985 — J. Searle, D. Vanderveken. Foundations of illo-cutionary logic. Cambridge, 1985.
Sharpe 2015 — D. Sharpe. Your chi-square test is statistically significant: now what? // Practical Assessment, Research & Evaluation 20, 8, 2015. P. 1-10.
Venables, Ripley 2002 — W. N. Venables, B. D. Ripley. Modern Applied Statistics with S. Fourth Edition. New York: Springer, 2002.
Walker et al. 1994 — D. Walker, Ch. Greenwood, B. Hart, J. Carta. Prediction of school outcomes based on early language production and socioeconomic factors // Child Development 65, 2, 1994. P. 606-621.
Watts 1986 — R. J. Watts. Generated or degenerate? // D. Kastovsky, A. J. Szwedek, B. Ploczinska, J. Fisiak. Linguistics Across Historical and Geographical Boundaries. Walter de Gruyter, 1986. P. 157-173.
