4.4. Propuesta de plan de mantenimiento preventivo
4.4.2. Plan de mantenimiento para scooptram diésel
Massengill Shaw, Carlson, and Waxman (2007) conducted an exploratory investigation into the relationship between texting and spelling with 86 American university students. These young adults completed questionnaires on their texting practices and standardized spelling tests. Massengill Shaw et al. found no significant correlation between texting frequency (number of messages sent per day) and spelling, neither perceived nor actual spelling ability.
Drouin and Davis (2009) studied the effect of textese on literacy with 80 American university students. Experimental methods were used to measure their textism use in different contexts (by writing formal vs. informal spontaneous emails in response to a scenario provided by the experimenter), textese proficiency (by translating Standard English sentences into textese), familiarity with textese (by translating textese sentences into Standard English), and misspellings of target words commonly abbreviated in textese such as you’re, to, two, and too (by recording spelling ‘errors’ for these words in translating into Standard English). Standardized tests assessed their reading and spelling skills. There were no significant differences between users and non-users of textese in their literacy scores or ‘misspellings’ of words regularly abbreviated in textese. Drouin and Davis’ findings are inconsistent with student perceptions: although many students believed texting to negatively affect literacy, no relations were found.
Spooren (2009) studied the relationship between online chat and writing quality with Dutch adolescents. 35 participants filled in questionnaires on their use of texting, IMing, and SNS and completed a writing task. The writings products were analysed offline at the global, lexical, grammatical, and textual level; the writing process was studied online too. Regression analyses were used to explore whether participants’ intensity of using CMC could predict the quality of their writings, yet this did not turn out to be the case. Spooren’s findings suggest that concerns about CMC affecting Dutch youths’ literacy skills are unnecessary.
Dürscheid, Wagner, and Brommer (2010) compared Swiss adolescents’ school writings to their out-of-school digital writings. They quantitatively analysed 953 school writings (essays) of secondary school students and 1148 of their CMC writings for writing features of textual coherence, lexis, morphosyntax, orthography, and typography. No impact of adolescents’ informal digital writing, which contained some salient CMC characteristics, was found on the essays. What is more, Dürscheid et al. found that an orientation towards the standard language in written CMC did not indicate a greater adherence to standard language norms in school texts. They also qualitatively analysed nine students’ writing portfolios, but these were neither indicative of interference from informal, digital to school writing.
Gann, Bartoszuk, and Anderson (2010) examined the association between texting and spelling ability, with 62 American university students and 44 adults from the surrounding community. Questionnaires were used to ascertain participants’ frequency of texting (number of messages sent daily) and use of textisms. A custom- made spelling test measured their spelling skills. Results showed no effects of texting frequency or reported use of textisms on spelling performance: the number of correctly spelled words did not significantly differ between those who texted or used textisms and those who did not. Gann et al.’s study found no evidence of associations between texting practices and adults’ spelling.
Radstake’s (2010) master thesis explored the relationship between spelling and new media use. Her participants were 352 Dutch adolescents at different levels of secondary education: lower secondary professional education (‘vmbo’), higher general secondary education (‘havo’), and pre-university education (‘vwo’). Participants completed questionnaires on new media (texting, IMing, email, and SNS), about the amount of time they used such media for social purposes, to keep in touch with family and peers. Their spelling ability was assessed with a standardized test. New media use did not significantly correlate with spelling ability and could not predict spelling ability in a regression analysis. Radstake thus found no relations between new media use and the spelling skills of Dutch adolescents from different educational levels.
To study the relationship between spelling and textism use in IMing, Varnhagen et al. (2010) conducted a naturalistic study with 40 Canadian adolescents, who were asked to collect all their actual IM chats for one week. A random one hundred-word sample from each participant was used for analysis. Participants completed a spelling test administered via an IM program: they typed words after having listened to recordings of single words and context sentences. Of course, all spell-checkers were disabled. Spelling ability turned out not to be related to textism use, but only to true spelling ‘errors’. This brings Varnhagen et al. to the conclusion that IMing does not affect adolescents’ Standard English spelling.
For his PhD thesis, Dixon (2011) investigated the relationship between youths’ intensity of using Facebook and their writing efficacy. 293 American students from a community college participated, among whom native and non-native speakers of English. Questionnaires determined their engagement with Facebook in two ways – amount of time spent on Facebook per day and number of Facebook friends. Their academic writing success was measured in three ways – self-reported writing
confidence, self-reported writing grades, and scores on writing samples. The samples, obtained from 189 students, were scored for content (“focus and development,” “introductions and conclusions”); organisation (“arrangement of paragraphs,” “transitions between ideas”); style (“language, tone,” “phrasing and sentence structure”); and mechanics (“grammar, punctuation and formatting”) (73). Neither measure of engagement with Facebook correlated significantly with any of the measures of writing success. The extent to which college students use Facebook apparently does not affect their academic writing success. Yet Dixon’s measures of engagement are somewhat flawed: the measure of time spent per day on Facebook does not distinguish between writing posts and playing games or reading others’ posts, and the measure of Facebook friends does not truly reflect intensity of use – one can have many such friends, for instance as a status symbol, without engaging with them via CMC.
In their study into the impact of texting on Pakistani students’ academic writing, Aziz et al. (2013) analysed the orthography of essays produced in class by 50 university students. The texts were examined for the occurrence of two ‘SMS features’ – texting abbreviations and omission of punctuation. Spelling was not affected at all: no textisms were found. Yet capitalisation and, in particular, punctuation marks were frequently omitted (or misused) in the essays, especially commas and full stops. However, the authors do not only attribute this to texting, but rightly suggest that other factors may be at play, such as students’ carelessness or inadequate knowledge of punctuation rules – possibly, in turn, due to insufficient training, feedback, or emphasis by teachers. Aziz and colleagues conclude that their study “has demystified the myth that SMS [has] disastrous effects on language in general and students’ writing proficiency in particular” (12890). In contrast, they argue that students revealed an awareness of different writing contexts, and an ability to switch between the informal register of texting and the formal register of academic writing. Nevertheless, punctuation remains a cause for concern, although the role of CMC herein is unclear.
Bernicot, Goumi, Bert‐Erboul, and Volckaert‐Legrier (2014) conducted a longitudinal intervention study in 2009–2010 with 49 French children, which lasted for twelve months, assessing the impact of texting on their writing and spelling skills. None of the participants had ever owned or used a mobile phone. 19 children, the experimental group, were given free access to mobile phones for the duration of the study (though quite old-fashioned ones, with alphanumeric keypads set in the multi- press mode of text input); the control group consisted of 30 other children who had no access to mobile phones. This method resembles that of Wood, Jackson, et al. (2011), but Bernicot et al.’s intervention lasted much longer (12 months vs. 10 weeks), increasing the chances of finding a significant impact of texting on the writing measures. They had two such measures. All participants took a standardized spelling test after nine months. In addition, non-standardized school writing grades were gathered before the data collection and twice during the collection. Even though the intervention lasted for a year, Bernicot et al. found no differences between the texters and non-texters in their writing or spelling performance at any of the testing moments. That children’s school writing skills were unaffected by their
text messaging might be because of the small sample size of the study. Still, the authors interpret these null results in a positive way and suggest that texting can be beneficial, by providing extra writing practice for children.
Besides analysing the attitudes of Canadian and Australian young adults on the appropriateness of textese in different contexts, Grace, Kemp, Martin, and Parrila (2015) investigated the intrusion of textese into the Australian youths’ formal school writings. 303 written exams of 153 university students were analysed for the presence of textisms. They found only very few textisms creeping into the exams – 117 textisms (0.02%) to a total of 533,500 words written, many of which (43) were used by only one student. Several textism types distinguished in previous studies did not occur in the exams at all, e.g. initialisms, combined homophones, accent stylisations, and extra capitalisation. This led Grace and colleagues to the reasonable conclusion that most students are not just able to evaluate in which contexts textisms are (in)appropriate, but are also able and willing to avoid them in school writings. Understandably, symbols such as “+” and “&” were not included in the analysis, since these can represent time-saving writing as a consequence of timed exam conditions rather than reflecting textese, yet it is a missed opportunity that Grace et al. did not count the number of misspellings in their corpus of exams, because this may have yielded more interesting results – although misspellings, obviously, cannot simply be attributed to the impact of CMC.
Rathje (2015) explored whether the use of new media affects Danish adolescents’ school writings. She compared 10 students’ text messages and Facebook messages, updates, and comments with their essays. Both were analysed for the occurrence of one type of orthographic reduction (a typical CMC language feature), namely verbal short forms, which may be used for reasons of economy or orality. The Facebook messages contained significantly more verbal reductions: in fact, the essays hardly contained any. Rathje thus did not find any relations between CMC use and the orthography of Danish adolescents’ formal writings. A point of criticism is that her analysis was quite restricted, since only verbal reductions were studied and no other textisms.
Sánchez-Moya and Cruz-Moya (2015b) conducted an exploratory experiment on the impact of WhatsApp on spelling, but they had a rather different approach. They appear to have been the first researchers to empirically investigate this issue with WhatsApp, yet their study was very limited in that they conducted no statistical analyses. Their participants were 15 Spanish adolescents and 15 Spanish adults. They were presented with a text that contained five traditional misspellings and five textese-driven ‘language errors’, such as missing capitalisation or extra punctuation. The adolescents turned out to identify more errors overall than the (highly educated) adults, especially regarding the errors based on textese. This suggests that youths may be more aware of register differences between CMC language and the standard language than adults. Sánchez-Moya and Cruz-Moya’s study thus gives no evidence of any relationship between CMC and literacy, since participants’ CMC use was not measured. Nevertheless, it is relevant in the debate about CMC and literacy, because they measured participants’ ability to correct textese-driven ‘language errors’ – a
method that is very similar to the grammaticality judgement task used in the experimental study of the present thesis (see chapter 10).
Finally, Ouellette and Michaud (2016) conducted a correlational study into Canadian young adults’ use of text messaging and textese and their language and literacy skills. They measured 51 university students’ textism use and ‘misuse’ of capitalization and punctuation (densities), by analysing ten of their recently sent naturalistic text messages and with a timed translation task, in which they had to translate Standard English sentences into textese. This task was also used to measure their speed of writing text messages, i.e. fluency with textese. Their texting frequency was gauged via self-reports. The students also completed standardized tests of spelling, reading, non-word reading (decoding), and vocabulary. Despite their thorough methodology, no significant correlations were found between any of the texting measures and any of the literacy skills. Ouellette and Michaud attribute this to the increased use of corrective and predictive technology in digital messaging. 2.3.5 Discussion of Observational Studies
An overview of the most important elements of earlier observational studies into the impact of CMC on literacy is presented in Table 2 in Appendix C, presented at the end of this thesis. The findings of the 50 studies discussed above, as presented in Figure 7 per age group, exhibit a decidedly mixed pattern of results: some found a positive association, some a negative association, others conflicting results (both positive and negative associations), and still others found no statistically significant association at all. This indicates that it is a highly complex issue in which various factors are at play. The inconsistent findings of prior research can, again, be contributed to the many differences in methodology and participants.
Figure 7. Findings of observational studies into the relations between CMC use and literacy. 12 8 16 14 0 2 4 6 8 10 12 14 16 18
Research design. As can be seen in Figure 8, an overwhelming majority of prior studies had a correlational research design (n = 37). Correlational or cross-sectional analyses are inherently limited, since they do not warrant conclusions about the causality between two variables. They cannot establish, when a positive or negative correlation was found, that this was a unidirectional effect of CMC on literacy. It may also be vice versa: a one- way effect of literacy on CMC, as suggested by Grace et al. (2013) and by Kemp (2010), who notes that in case of a positive association, youths with better literacy skills may “better employ these strengths to create and decipher textisms” than those with weaker
literacy skills (65). The effects between CMC and literacy may also be reciprocal: for instance, Durkin et al. (2011) hypothesize that the positive relationship they found is bidirectional, that better literacy skills affect the ability to use textisms and that more frequent use of CMC at the same time helps develop literacy skills. Another option is that there are cognitive factors at work, such as participants’ IQ or verbal ability – what Kemp (2010) calls “an underlying level of linguistic or general intelligence” (65): a correlation does not necessarily imply causation, because a third variable may account for it. In correlational studies, all this remains conjecturing.
Only a number of prior studies have looked into the direction of the relationship: a pure experiment (Powell & Dixon, 2011), an exploratory experiment (Sánchez-Moya & Cruz-Moya, 2015b), two experimental studies using intervention (Wood, Jackson et al., 2011; Bernicot et al., 2014), and five studies with a longitudinal design (Wood, Meachem et al., 2011; Wood, Kemp, & Waldron, 2014; Waldron, Wood, & Kemp, 2016; Bernicot et al., 2014; Simoës-Perlant et al., 2018). The experimental studies indicate that textisms may affect literacy rather than the other way around, but again their results were mixed.
Six prior observational studies were corpus studies, which analysed school or academic writings for the presence of textisms (Winzker, Southwood, & Huddlestone, 2009; Dürscheid, Wagner, & Brommer, 2010; Shafie, Azida Darus, & Osman, 2010; Rankin, 2011; Grace et al., 2015; Vandekerckhove & Sandra, 2016). Corpus analyses reveal interesting but, again, limited insight into the relation between CMC and literacy, since they show to what extent non-standard orthography is present in formal writing, but cannot conclusively prove that these orthographic deviations were caused by CMC.
Future observational studies should, ideally, be of an experimental nature, because longitudinal studies with a control group of non-users of CMC are unfeasible
Figure 8. Research design.
37 6 5 2 2 correlational corpus longitudinal intervention experiment
in this digital age in which practically all youths are heavy users of social media. The present thesis tackles this challenge by conducting both a correlational study (chapter 9) and a study with experimental intervention (chapter 10).
Medium. Although an
overwhelming number of the studies focused on text messaging (46), as Figure 9 shows, there was still some diversity in the media that were included in prior observational studies. Instant messaging was included in 12 studies. Emailing (5) and social networking sites (4) were both studied to a much lesser extent. Some CMC modes have barely been investigated in relation to literacy at all: Facebook was included in only three studies and was the focus of two of those (Dixon, 2011; Rathje, 2015), while IMing with the currently very popular mobile phone application WhatsApp Messenger (3) has only been considered in a handful of recent studies in Spain, Belgium, and the Netherlands (Sánchez-Moya & Cruz-Moya, 2015b; Vandekerckhove & Sandra, 2016; Van Dijk et al., 2016). Furthermore, Bouillaud, Chanquoy, and Gombert (2017) mention online forums, and Vandekerckhove and Sandra (2016) use the term ‘chat language’, which may include texting, IMing, WhatsApp, and Facebook chat. One study focused specifically on predictive texting (Waldron, Wood, & Kemp, 2016), i.e. using predictive software. Blogging and websites were only included by Dürscheid, Wagner, and Brommer (2010); microblogging platforms such as Twitter were not mentioned at all. It is a clear limitation of prior observational research that a large majority focuses only on texting. Ideally, observational studies should include a range of social media, since these may have diverging effects on youths’ literacy skills: we cannot simply pool different media together in terms of their impact on literacy. This thesis overcomes this limitation by including various social media, both in the studies on language use in Dutch youths’ written CMC (presented in part 1, chapters 5–7) and in those on the relations between their CMC and school writings (part 2, chapters 8–10). The corpus studies in chapters 5, 6, 7, and 8 examine two CMC modes that used to be popular in the Netherlands (MSN chat, SMS) and two that are still popular (Twitter, WhatsApp); the correlational study in chapter 9 includes many CMC modes in a long survey; and the experimental study in chapter 10 focuses on the currently most used CMC mode by Dutch youths, WhatsApp (Van der Veer et al., 2018).
Operationalization of literacy. The studies greatly differed in the way in which they operationalized literacy (see Figure 10). Many studies employed direct
Figure 9. Medium. 46 12 5 4 3 3 1 1 1 texting IMing emailing SNS Facebook WhatsApp forums (micro)blogging websites
measures of literacy, measuring spelling, grammar, writing, or reading. The most frequently measured literacy skill was spelling (accuracy), which was included in 41 of the 50 studies discussed here. Reading measures were used seventeen times, including fluency/efficiency/ speed, accuracy/word recognition, comprehension, history, and novel word reading. Measures of writing (quality, efficacy, speed, accuracy, expressiveness/conciseness) occurred twelve times, those of grammar (accuracy, understanding, ‘grammatical spelling’) nine times. Besides such direct operationalizations of literacy, various indirect operationalizations were also used, measuring integral components of literacy – skills underpinning literacy skills. This includes measures of non‐ word reading or decoding (alphabetic/
orthographic/phonological (9));
awareness (phonological awareness/skill
(7), orthographic awareness (1), morphological awareness (2)); processing (phonological processing (1), orthographic processing (4)); retrieval (phonological retrieval (3), lexical retrieval (1)); and reasoning (verbal reasoning (1), non-verbal reasoning (2)). Cognitive skills (vocabulary (5), short-term memory (2), rapid serial naming (1), executive functions (1)) were used as literacy-related measures or controlled for as confounding variables (Plester, Wood, & Joshi, 2009; Wood et al., 2014). One study did not measure language ability, but rather language acceptability. Although these indirect measures are relevant for literacy, any impact of CMC on such measures is less probable than on direct literacy skills. In the present thesis, literacy is thus operationalized in terms of a direct measure, namely writing quality.