Deepening the Analysis of Literary Texts among University Students Using Close Reading and Writing: A Pilot Study

António M. Duarte1, Cecilia Beecher Martins2, Maria C. Marques3 and Filipe C. Mesquita3 1 Centro de Investigação em Ciência Psicológica, Faculdade de Psicologia, Universidade de Lisboa, PT 2 Centro de Estudos Anglísticos, Universidade de Lisboa, Faculdade de Letras, Universidade de Lisboa, PT 3 Faculdade de Psicologia, Universidade de Lisboa, PT Corresponding author: António M. Duarte (amduarte@psicologia.ulisboa.pt)

This essay offers an analysis of the findings derived from a pedagogical case study performed with university students. This study set out to evaluate the effects training in close reading and close writing, applied to literary fictional texts, had on the structural complexity of narrative analysis (SCNA) of participating students. Participants' perceptions of the impact of this training on their reading and writing skills was also monitored. SCNA was measured using Biggs and Collis (1982) SOLO taxonomy and qualitative thematic analysis was employed to analyse a focus-group interview on the impact of the training. The sample was randomly divided into two equivalent groups, one experimental (EG), the other control (CG). Both groups did three evaluations to access SCNA at the same time intervals: pre-training, post-training and follow-up timings of the EG. The CG only received training after they completing all three evaluations. While pre-training results demonstrated similarity across groups at the outset of trial after training, the EG's SCNA increased significantly from pre-test to post-test and became significantly higher than the CG's, which did not differ significantly. The EG's gains were maintained in follow-up. Finally, the interview demonstrated that participants in the EG perceived that the training improved their reading and writing skills. Keywords: close reading; close writing; narrative analysis; reading skills; writing skills Before presenting the pilot study, we will introduce the concepts of close reading and close writing in educational practices today. 1 Close reading (an analysis focusing on the structures and patterns present in the text), is recognized to have its roots and continued practice in the theological process of exegesis (the interpretation of sacred texts) and in literary criticism. In these areas it is used as a tool of enquiry and scholarship. Brookman and Horn's recent pilot study, using close reading to analyse students' abstracts and evaluation forms, speaks to this continued use (248)(249)(250)(251)(252)(253)(254)(255)(256)(257)(258)(259)(260)(261)(262)(263)(264)(265). However, today much of the research on close reading looks at it as a tool to enhance functional literacy (Sisson and Sisson, The Renaissance). While the work we carried in our pilot study fits into the latter category of functional literacy, we realise that it was informed by the former. Furthermore, our approach to close reading and close writing-producing practical analysis and expression strategies-was also influenced by educational psychology. Thus, all three perspectives on close reading are presented in the conceptual framework.
Moreover, within the context of functional literacy, Sisson and Sisson propose that close reading could be coupled with close writing. They suggest that when close writing is carried out sequentially after close reading and involves writing a creative piece, based on the previously close read text, this close writing activity will synthesize the understanding of the text read because "If students are to learn, they must be able to express their understanding through their own writing" (Close Reading 170). In Active Reading (2008), Knights and Thurgar-Dawson also couple active reading and transformative writing (arising from the active reading) for the teaching of English literature in a third level education context. They go so far as to suggest that these reading and writing practices can assist lifelong learning and professional development, Imagination, disciplined through transformative writing, may be (to put it reductively) an implement of professional development. But if so, its power rests on the capacity of what we might call co-writing, or writing alongside the text to activate simultaneous levels of knowledge and of the self. (150) This coupling of attentive/close reading and writing for lifelong learning and professional development is found in other education and professional contexts, namely in narrative medicine training and practice. Narrative medicine proposes that training health care professionals in the close reading of literary narratives can help professional careers develop a more humanistic approach to caring because it facilitates active or close listening (Charon, Narrative Medicine). In fact, Charon and her team classified close reading as the "signature method" of narrative medicine because it assists healthcare professionals in the development of the professional and interpersonal skills they need to function well in all aspects of, and throughout, their professional lives, because it "reflects and articulates the foundational principles of narrative medicine" (Charon et al.,The Principles 8). Moreover, in narrative medicine training close reading is followed by creative writing. This writing activity is carried out in "the shadow" of the text read, "to encourage and develop within learners the skills to represent complex events or states of affairs so that they can be perceived and maybe understood" (Charon et al.,Close Reading 345). Thus, parallels can be seen between close reading and writing applied in general education and close reading and creative writing actives of narrative medicine.
While many proposals are made about the benefits of close reading and writing, we are unaware of any study on how training in these skills can improve functional literacy among university students. The only studies we found- Dakin (2013) and Glover (2017)-were carried out in primary education. Thus, we decided to carry out a pedagogical case study to gain some preliminary insight into this subject. As we were operating in terms of functional literacy, we wanted to work with students coming from different academic backgrounds, not just the humanities. Thus, we elected to use the close reading and close writing format used in narrative medicine workshops, as this would appeal to both healthcare and humanities students. Moreover, it is appropriate for the age and academic background of university students. We selected the Biggs and Collis (1982) SOLO taxonomy to measure alterations in the structural complexity of narrative analysis as this is a recognised tool and often used to measure the structural complexity of narrative analysis. The SOLO taxonomy would be supported by a focus-group interview to see if/how the close reading and writing activities influenced participants' perspectives of how the intervention changed their ability to read and write on literary texts.
We will present the outcomes of this pilot study in this essay. However, before presenting the study we will look at how close reading is defined in the fields of functional literacy, literary analysis, and modern educational psychology because, as mentioned above, the practical work was informed by all three.

Close reading and functional literacy
The development of state-led efforts to harmonize, not only teaching standards, but also expected outcomes for students at a national level (for example the Common Core State Standards-CCSS-in the U.S. and the Quality Assurance Agency-QAA-in the U.K.) as well as the rise of benchmarking by internationally recognized vetting organizations (for example, the Programme for International Student Assessment-PISA) has meant that teaching tools offering functional capacities across cultures have been sought. Close reading has been seen as one such tool.
The authors of the Common Core State Standards for English Language Arts & Literacy in History, Social Studies, Science, and Technical Subjects define close reading as the ability to "interpret words and phrases as they are used in a text, including determining technical, connotative, and figurative meanings, and analyse how specific word choices shape meaning or tone" (10, 35). They suggest that close reading is not only an education tool, but also an instrument that can facilitate the development of critical and analytical skills, in other words the functional literacy necessary to function well in contemporary society, because they suggest, Students who meet the Standards readily undertake the close, attentive reading that is at the heart of understanding and enjoying complex works of literature. They habitually perform the critical reading necessary to pick carefully through the staggering amount of information available today in print and digitally. They actively seek the wide, deep, and thoughtful engagement with highquality literary and informational texts that builds knowledge, enlarges experience, and broadens worldviews. They reflexively demonstrate the cogent reasoning and use of evidence that is essential to both private deliberation and responsible citizenship in a democratic republic. (3) Because of the pivotal role ascribed to close reading, the authors suggest that training in this activity should begin at the very outset of formal education: "Preparation for reading complex informational texts should begin at the very earliest elementary school grades" (33) because the authors consider the capacity to "read (age appropriate texts) closely to determine what the text says explicitly and to make logical inferences from it" (10,35), should be a central pillar of education.
However, while the above offers indications, both of what close reading is and of expectations of what it can produce, the proposals are very broad and sweeping. As Brookman and Horn have pointed out, forms of interpretation associated with close reading practices have become, "a large, eclectic and contested field" (251). Therefore, we considered it necessary to return to the original proposals of what could be expected from close reading in the context of literary analysis to provide a background for our pilot study. Particularly, we aspired to ascertain how its original use as a tool of educational enquiry and scholarship might inform and propel its usage as a tool of functional literacy in the context of a narrative medicine style workshop.

The tradition of close reading in literary criticism
In Literary Theory: An Introduction (1983) Terry Eagleton suggests that it was A.I. Richards' work-connecting literary studies on both sides of the Atlantic in a formalistic approach to textual analysis-that brought close reading to the fore in literary studies (Eagleton,. Most scholars agree that the use of close reading, as a formal tool of enquiry in literary criticism, dates back to two of A.I. Richards publications: The Principles of Literary Criticism (1924) andPractical Criticism (1929). While John Crowe Ransom would expand on Richards' theories in The New Criticism (1941) and this volume would go on to name the fledgling Anglo-American literary analysis movement that placed close reading at its core, the initial application in literary scholarship began with Richards. Because of the foundational position of his work in the field, we will begin this section by examining Richards' presentation of close reading and progress to look at its application in literary studies since its introduction.
In The Principles of Literary Criticism Richards suggested that interpretation of poetry should not be dependent on understanding the poet's conscious or unconscious intent when writing the poem (psychoanalytical theories of interpretation were in vogue at the time Richards was working and writing), or even mastering all the technical aspects of rhythm and meter. Rather, he proposed that meaning in poetry is conveyed through the poet's individual use of language, and the reader can understand and interpret a poem by paying close attention to the language and the formalistic structures and patterns presented in the language of the poem. He suggests that poetry (and the arts in general) had been misunderstood by both formal and informal critics because not enough attention had been given to the examination of the formal structure of the object itself, Often those who most misunderstand have been perfect in their taste and ability to respond, Ruskin for example. (…) Those who have tried have as a rule been foiled by language. For the difficulty which has always prevented the arts from being explained as well as "enjoyed" (to use an inadequate word in default of an adequate) is language. (27) Thus Richards proposes that concentration on the language of literature, the words and form of expression, will not only lead to understanding, but also enjoyment.
Perhaps for this emphasis both on enjoyment and understanding, Richards' work caught the imagination of many of his more gifted students and writers of his day, including William Empson and T. S. Eliot, who not only promoted it, but also expanded it (Eagleton,Literary Theory 40) and New Criticism would remain popular in the United States until the 1950s (79). But even when the sway of New Criticism waned; its central tenet-attention to structure and form-remained in the subsequent structuralism movement, which Eagleton proposed "brackets off the actual content of the story and concentrates entirely on the form" (83). The 1970s saw the rise of different interpretative systems, from reader response to psychoanalytical literary criticism among others; still close reading was retained in English literature departments. As Jonathan Culler wrote in his 2012 essay "The Closeness of Close Reading", In many English departments, and I daresay foreign language departments as well, the practice of close reading, of examining closely the language of a literary work or a section of it, has been something we take for granted, as a sine qua non of literary study, a skill that we expect our students to master and that we certainly expect of job candidates, whatever other sorts of critical activities they may flamboyantly display. (20) These references speak to the continued use of close reading in literary analysis to this day. Moreover, when this skill is not present, it is missed, as Terry Eagleton pointed out in How to Read a Poem (2007), where he complained about students' lack of training in close reading and their tendency to jump to conclusions about issues of gender, race or class, without paying due attention to the language of the text (1)(2)(3)(4)(5)(6)(7)(8)(9)(10)(11)(12)(13)(14)(15)(16).
This statement of Eagleton's is particularly interesting because his initial response to Richards's work had not been particularly favourable. Eagleton's Marxist approach to literary analysis demanded the consideration of the historical context for the understanding of a literary text, which did not agree with Richards' insistence on attentiveness to the text, "to the 'words on the page' rather than to the contexts which produced and surround them" (Literary Theory 38). Eagleton proposed close reading was not as neutral as Richards' advocated because when reading Richards' students' analysis, Eagleton suggested that it was easy to see that their interpretations had been influenced by their socio-cultural background and their expectations of what a poem should be. Eagleton proposes that Richards had not seen this because he shared both elements (13). However, even when writing Literary Theory, Eagleton also appreciated that when students analysed texts following Richards principles, they "were highly variable: time-honoured poets were marked down and obscure authors celebrated" (13), so Eagleton also recognised uniqueness and individuality in students' analysis.
Moreover, in recent publication we can see scholars call for the use of practices that bear striking similarities to close reading, even if they do not use that particular term. For instance, in A Future for Criticism (2011), Catherine Belsey argues that literary criticism should "concentrate on language in general and textuality in particular" (85). Moreover, even if Derek Attridge and Henry Staten suggest "'close reading' is too ideologically radioactive, and means too many different things" (2) in The Craft of Poetry: Dialogues in Minimal Interpretation (2015) the concept of minimal interpretation they introduce corresponds with close reading as it also revolves around a detailed and disciplined analysis of the words in the text before application of other analysis methods (2).
Thus, we can conclude that when students analyse texts by paying close attention to the form and structure of the language, they can engage with and enjoy literature as they gain a deep and personal understanding of the text read. For this reason, while this tool may have been developed initially for educational enquiry and scholarship, it is also relevant for functional literacy, lifelong learning and personal development as it allows readers to develop a self-referential capacity of analysis, which can be applied in all three areas.

Close reading as defined in modern educational psychology
As the goal of this pilot study was to experimentally test the impact of close reading and close writing on the practical analysis and expression strategies of students in higher education, concepts linked with educational psychology, we deemed it necessary to look at how close reading and writing are considered by this field also. On examination we discovered parallels can be seen between the definitions and applications of close reading in functional literacy, literary analysis and educational psychology, with educational psychology fine tuning more general proposals coming from literary criticism, as well as expanding its reach to non-literary texts as functional literacy also does.
Elder and Paul defined close reading as a reading strategy that involves "moving closer" to the text (literary or informative), contrasting it with "superficial uncritical reading" (289). Beers and Probst suggest that close reading requires a deliberate high attentional examination of the elements of the text studied (2013), which in turn permits the analysis of deep structures (Fisher and Frey) and meaning (Brummett).
Close reading has been defined by a number of authors in this area as involving a diversity of complementary reading operations: determining the level of importance of each idea of the text (Cummins); defining its key elements (Fisher and Frey); analysing how ideas relate (Cummins); identifying how the text is organized (Fisher and Frey); relating the text with one's own background knowledge (Brown and Kappes; Dakin); identifying the arguments, messages or meanings delivered by the text and the evidence on which they are based (Boyles; Cummins; Fisher and Frey); interpreting the text (Sisson and Sisson, The Renaissance); and/or developing a critique of its ideas (Cummins). It is important to note that while we acknowledge all of the aforementioned may be used in the close reading of any given text, they do not all need to be used in the same application. Moreover, we recognise, as suggested by Fisher and Frey, that close reading may also require that the above-mentioned cognitive operations are supported by the annotation of the text in a variety of ways, like underlining, circling or writing notes in the margins. Curiously, this is a practice that Charon et al. also advocate in narrative medicine literary analysis workshops.
Some auxiliary conditions have been specified to facilitate the development of close reading and writing skills. Various authors (Beers and Probst; Boyles; Cummins; Fisher and Frey) suggest that close reading can be developed by answering questions that bring together the operations required to perform this skill (for example asking a question like "What is the message of the text?"), by repeating readings of the same text (Fisher, Frey and Lapp, 2012) and by discussion among small groups of students on individual answers to questions of the type mentioned above (Brown and Capes; Sisson and Sisson, Close Reading). Moreover, Beers and Probst suggest that providing students with background information on the content of the texts may also be necessary to allow the student to relate to the text and comprehend it. There is some discussion about the optimal length of the texts for use in this type of activity: Fisher and Frey refer to the importance of using short texts, to optimise the practice of close reading, while both Dakin and Sisson and Sisson (Close Reading) suggest that texts should be complex and challenging independent of length.
A number of advantages have been proposed for the use of close reading and close writing in opposition to "superficial reading and writing". Elder and Paul found parallels between close reading and close writing when they compared these two activities. They suggest that as close reading contrasts with "superficial uncritical reading", close (or substantive) writing contrasts with "superficial writing", which they define as writing that does not state anything substantial (289). They imply that close reading helps students to better retain a more precise image of what they have read. The same authors indicate that substantial writing helps students to better comprehend and take ownership of what they write, suggesting that students' inadvertent use of "superficial writing" prevents them from effectively understanding and appropriating for themselves a true appreciation of what they have written. Superficial reading is seen as leading students to forget and/or distort what they read. Also, Dakin claims that close reading promotes understanding and critique of the text, accelerating and deepening learning, thus contributing to the development of independent learners and thinkers. Dakin also refers to the adaptive values that close reading may offer in a world where readers are continuously exposed to information and need to comprehend in order to be active in their community.
The suggestions above echo the proposals made for close reading and close writing in literary analysis and functional literacy. In fact, we included some of the techniques mentioned above in the close reading and close writing workshops of the intervention, and thus the incursion into educational psychology enriched the work. The proposals also fit with Charon et al.'s proposals (Close Reading).

The goals of the study
The goal of this study was to experimentally evaluate the effects of the use of close reading and close writing (intervention) on Humanities students (HS) and Medicine students (MS). Two effects will be measured: the quality of analysis of literary narratives on the thematics of health (measured using SOLO taxonomy) and perceived intervention induced changes in reading of and writing on literary narratives (measured by a focus-group interview). In this context close reading and close writing were used as purveyers of practical analysis and expression strategies rather than tools of educational scholarship, even if their use was informed by the latter. In this sense, the individual engagement with and enjoyment of the text would be encouraged, as would the development of a self-referential capacity of literary analysis. However, these would be seen as practical analysis and expression strategies not as educational tools in a formal academic sense.

The hypothesis of the study
In this study three hypotheses were established to access the impact of an intervention (that used close reading of literary texts and close writing based on this reading) on the quality of analysis of narratives. H1contemplated that after the intervention there would be a general increase in the structural complexity of narrative analysis (SCNA) among participants of an experimental group illustrated by the SOLO taxonomy results. H2 contemplated that the changes predicted in H1 would be maintained in a follow-up examina-tion of participants in the experimental group. H3 contemplated that despite the general changes predicted in the experimental group in H1 and H2 there would be interindividual variability within this group, with only some cases benefiting from the intervention. Biggs and Collis (1982) designed the SOLO (Structure of the Observed Learning Outcome) taxonomy, which distinguishes between preoperational, concrete and formal thinking in the analysis that individuals perform of any content. Biggs and Collis (1982) propose that an analysis of any content involves three parameters: "relating" (i.e., logical connection with the content); "capacity" (i.e., amount of information presented); "consistency and closure" (i.e., how the analysis concludes).

Presentation of the Biggs and Collis SOLO Taxonomy (1982)
Through the observation of how these dimensions vary in different learning products, the authors developed a taxonomy that characterizes five levels of increasing complexity: 1) pre-structural (the lack of a logical relationship between the analysis and the content under analysis, the analysis is repetitive and/or irrelevant and inconsistent with the content); 2) uni-structural (the analysis is logically associated with the related content, including one correct and relevant informational unit); 3) multi-structural (the analysis is logically associated with the content with some relevant but not interconnected informational units that itemise or replicate content); 4) relational (the analysis is logically related to the content with quite a few pertinent integrated informational units in the structure of a concept or argument); 5) abstract (the analysis is logically connected to the content and also possibly considers hypothetical questions, including interconnected information and general principles which form abstract connections with other content domains). The SOLO taxonomy can be applied to the evaluation of analysis in different knowledge areas, including literature, and has been used in the research of the connection between the learning process and learning products (Biggs and Tang). Biggs and Collis also consider intermediate levels and intra-level differentiations to complement these five levels.

Method & Procedure
The impact of the intervention-a workshop training the small experimental group in close reading of literary texts and close writing to the same-(independent variable) on participants' SCNA (dependent variable) was assessed to test H1 following a before-after experimental design (Christhensen). The SCNA of the experimental group and the control group was evaluated before (pre-test) and after the intervention (post-test) and in a follow-up (H2). In addition, on the basis of a multiple case experimental design (Cristhensen, Neuman and McCormick, Wilson) a before-after intervention contrast of SCNA was performed for each participant in the experimental group to establish if the effects of the intervention were specific to each individual (H3). To complement this, participants' SCNA pre-intervention results were compared with SCNA follow-up results. Finally, after intervention, a qualitative analysis was performed to determine the perceptions that participants in the experimental group had on the impact of the intervention on their reading and writing.

Participants
The above-mentioned workshop was made available at the University of Lisbon campus to medical and humanities students (open enrolment). Four students from a Medical faculty and fourteen from a Humanities faculty enrolled. They were randomly assigned to the experimental and control group, after predetermining that each group would be composed of two medical students (MS) and seven humanities students (HS). The experimental group was composed of nine students: two males and seven females, mostly in the third year of their studies (M = 3; SD = .44) with a statistical mean age of twenty-two point five years (SD = 3.68). The control group was also composed of nine students: two males and seven females, mostly in the third year of their studies (M = 3; SD = 1.36) with a statistical mean age of twenty-two point eight years (SD = 5.51). No participants dropped out of the study. The experimental and control groups were equivalent regarding their general SCNA before intervention (i.e., there were no statistically significant differences between the SCNAs of these groups). Informed consent was obtained after participants were acquainted with the procedure, context and confidentiality of the study. Ethical approval for this study was granted by an academic institution's deontological commission.

Measures and Measurement
All participants, both in the experimental and control groups, read excerpts of literary narratives (without guidance or instruction) at three time points: pre-intervention (3 extracts), post-intervention (3 extracts) and follow-up (1 extract). After reading each text participants answered the question "what is this text about" in writing. To obtain a SCNA characterization these written answers were categorized according the SOLO tax-onomy proposed by Biggs and Collis. Half-point scores were used considering the intermediate levels these authors propose. The participants' texts were classified by two independent analysts comparing participants' written answers with examples constructed for this study. If disagreement existed, this was resolved through a discussion and/or an appeal to a third judge. The percentage of agreement (PA) between the analysts was eighty-eight-point nine percent (PA = (agreements ÷ agreements + disagreements) ×100). SCNA results were presented as mean values. The examples were "blindly" validated by an evaluator with experience in the SOLO taxonomy. Examples of answers for the five levels of SOLO taxonomy for one of the texts used in the evaluation (extract from Lydia Cassatt Reading the Morning Paper by Harriet Scott Chessman) can be seen in Appendix 1.
As we would be working with a heterogeneous group of medical and humanities students, we opted to use extracts from literary prose dealing with illness. The texts used in pre-and post-evaluation were taken from the anthology Contar (com) a Medicina (2015) A complementary qualitative assessment of the experimental group's perception of alterations in reading and writing skills as prompted by the intervention was carried out through a semi-structured focus-group interview. The interview was implemented using a prepared script that covered participants' general perception of the intervention, questioned general changes eventually provoked by it and enquired into participants' perception of changes in their capacity to read and write on narratives because of the intervention.

Intervention -Close Reading and Close Writing Workshops
Members of the experimental group took part in three fifty-five-minute close reading and close writing workshops carried out over a period of three weeks. The participants practiced close reading and writing with excerpts of literary narratives from the following novels: The Whereabouts of Eneas McNulty (2008) by Sebastian Barry, Heart and Soul (2008) by Maeve Binchy, and The English Patient (1992) by Michael Ondaatje. These texts were chosen because like the texts used in the evaluation they deal with different aspects of care and illness. The first was an observation of illness from the outside; the protagonist sees his friend suffer an epileptic fit for the first time. The second, offers a description of illness as it is experienced; the protagonist suffers a heart attack and describes what he feels physically and emotionally as he suffers the sympthoms and these are diagnosed. The third describes a situation of care; a nurse takes care of a patient suffering from severe burn injury.
Following the guidelines of narrative medicine workshops (and educational psychology), in each session participants were given a brief introduction to the text. This was then read aloud, slowly and expressively, as participants read it silently and freely marked (i.e., underlining and/or making notes) the aspects they considered as relevant. Subsequently, each participant was invited to speak about sections they had highlighted as relevant to them and these were discussed to assist comprehension and consolidate personal connections with the text. As this process developed, the facilitator called participants' attention to the literary techniques and language used in the text (e.g., metaphor, irony), to illustrate how these devices had attracted the reader's attention to specific elements-both those expressed and those implied-in the text.
Close reading was followed by a five-minute close writing exercise (this could follow an open or constructed form of prose or poetry) written to a prompt related to the text read (e.g., the prompt used with Maeve Binchy's text was "Life as they knew it was over"). Subsequently, participants were free to read their writings aloud to the rest of the group, and written texts were discussed using the close reading technique used in the analysis of the original literary text read. It is relevant to mention here that the role of leader of a narrative medicine workshop is that of a facilitator, who tries to help participants understand their individual connection to the text as well as to develop individual analytical and expression skills, rather than a teacher who wishes to convey particular knowledge about the text.
Although the control group did not participate in the intervention, as participants were placed in the control and experimental groups arbitrarily and many participants in the control group were interested in doing the workshop, they were offered the same workshop after the follow-up evaluations were completed. Four members of the control group did this workshop and transmitted orally to the facilitator that they had enjoyed the experience. 2

Data Analysis
First, the equivalence between the experimental and the control group in terms of their SCNAs before intervention was ascertained using the U Mann-Whitney test. Second, H1 and H2 were also accessed using the U Mann-Whitney test. Then the Friedman's 2-way ANOVA by ranks nonparametric test with 3 related samples was also applied to each group, taking the dimension of the sample into consideration. As indicated above, SCNA was assessed at the three time points: pre-test, post-test and at follow-up. Besides applying this statistical test to the data of both groups, the Wilcoxon's signed rank test was performed for post-hoc analysis in order to evaluate changes in users' SCNA. Third, the effect of the intervention was verified for each individual case by comparing the measurements of individual SCNA before and after the intervention to test H3.
Finally, a thematic analysis (Miles and Huberman) on the perceptions of intervention's impact on reading and writing was carried out to analyse participants' answers to the interview performed to evaluate this point. The answers to the interview were first deductively segmented into units on the basis of a thematic criterion (changes in reading or in writing). Resulting units were then inductively categorized, in order to construct a categorizing system that was then applied for a second loop categorization of all the units. Validity of the categorization system was studied through its application across all of the segmented units by an independent judge and calculus of agreement with categorization (agreement coefficient = 97.1%). The formula suggested by Bakeman and Gottman (1986) and previously presented (in Measures and Measurement section) was employed for the calculus.

Results
Regarding the impact of the intervention on SCNA ( Table 1; Graph 1) the experimental group (Mean Rank = 12.33) is statistically significantly higher than the control group after intervention (Mean Rank = 6.67), U = 15.00, p = .02, r = -.61 and significantly different at the three evaluation time points (χ²F (2) = 6.91, p =.034), vindicating what had been hypothesized in H1. SCNA of the experimental group increased significantly (p < .05, r = -.64) from pre-test (Mean Rank = 1.44) to post-test (Mean Rank = 2.33). In contrast, SCNA of the control group did not differ significantly (χ²F (2) = 3.71, p =.17) at the three time points, although it increased slightly from pre-test (Mean Rank = 1.56) to post-test (Mean Rank = 2.11) and from this to followup (Mean Rank = 2.33).
To determine if the intervention-induced changes would be maintained even in the period when there was no training (H2), pairwise comparisons were carried out to identify significant differences in the experimental group. This test disclosed that SCNA was not statistically significantly different between follow-up (Mean Rank = 2.22) and post-test (Mean Rank = 2.11), thus the alterations obtained during the period of the intervention were maintained for the month when no training was undertaken. Moreover, post-test (Mean Rank = 2.33) was statistically significantly higher (p < .05, r = -.60) at follow-up (Mean Rank = 2.22) when compared to pre-test (Mean Rank = 1.44).
The individual results were then compared before and after the intervention to determine the interindividual variability of intervention-induced changes within the experimental group. (H3). This showed that SCNA increased in 6 cases, was maintained in 2 cases and decreased in 1 case ( Table 2; Graph 2). It must also be noted that in the cases where SCNA increased with the intervention, when pre-intervention 2 All the members of the control group had demonstrated interest in participating in this optional workshop at the outset of the pilot study. However, because this occupied most of the second semester, the workshop for the control group had to be scheduled for very late in the academic year when students were preparing for final examinations or writing internship reports, so their availability was reduced.  =: equal to before the intervention. >: higher than before the intervention. <: lower than before the intervention.
Graph 2: SCNA of participants in the experimental group.
results were compared with follow-up results, the increase obtained in post-intervention was maintained in follow-up in all cases. The qualitative analysis of the interview with the experimental group revealed a perception that the intervention induced alterations in the way participants read narratives and wrote about them. In fact, substantial changes were registered in terms of how participants perceived their interaction/capacity to interact with literary texts. As can be seen in Table 3, a number of interesting phenomena were registered including: an increased awareness of the practices of reading; more motivation to read; and an improvement in a variety of cognitive processes associated with reading (i.e., attention, analysis, comprehension and reflection on text). More modest changes were registered in the field of close writing on narratives read. Still, some were registered and these included an increased awareness of the (deficit in) practices of writing and a general improvement in writing. Table 3 summarizes these results and each category is illustrated with an extract from the interview.

Conclusions
The experimental group's SCNA results confirm H1, which predicted a general increase of SCNA after intervention among participants in that group, as measured by the SOLO taxonomy. This indicates that when university students practice close reading of and close writing on literary narratives, the quality of the analysis they are able to carry out on this type of narrative can increase. This supports the proposal that the practice of close reading and writing is associated with a higher quality of textual analysis (e.g., Sisson and Sisson Close Reading). Considering that close reading and writing encourages a deeper analysis of content, the result aligns with SAL (Student Approaches to Learning) theory, which furnishes evidence that a deep approach to learning (i.e. the combination of intrinsic motivation to learn and a comprehension learning strategy) can be related with learning results at higher levels of the SOLO taxonomy (Trigwell and Prosser).
In parallel, the increase of SCNA (although statistically non-significant) of the control group from pre-test to post-test to follow-up can be interpreted as a sign that the recurrent active reading of and writing on literary narratives (i.e., in the assessment tasks) contributed to some (but statistically non-significant) increase of SCNA, which developed gradually from pre-test to follow-up. Moreover, the question participants had to answer on the texts in the pre-test, post-test and follow-up evaluations "What is this text about?" require the reflection some authors in educational psychology (Beers and Probst; Boyles; Cummins; Fisher and Frey) associate with close reading. Thus, it could be proposed that the control group's active reading and writing had some of the characteristics of close reading. However, it must be noted that the increase in SCNA was achieved much more quickly in the experimental group so the active training of close reading and close writing skills brings benefits. The experimental group's SCNA results also confirm H2, which previewed that the changes predicted in H1 for the experimental group would be maintained in follow-up. This maintenance suggests that close reading not only initiated but also stabilized increases in the quality of the analysis of narratives. Also, the individual results confirm H3, which predicted interindividual variability within the experimental group. The individual results reveal that in the majority of cases, the intervention led to a stable increase of SCNA, which confirms the generally positive effect of the intervention.
A minority of cases not following this pattern was noted, revealing that for these participants, the intervention had a null or even negative effect on SCNA. This suggests that the intervention was not sufficient to increase the SCNA of some participants, maybe due to the length of the interventions (it was quite short), structure of the format and/or personal characteristics. For instance, individuals whose "visual-spatial intelligence" is higher than the "verbal-linguistic intelligence" (Gardner, Multiple Intelligences 2006) might gain less from an analysis of literary narratives than of visual narratives. Moreover, some individuals might need a more extended practice of close reading than the one used in the intervention to benefit significantly from it.
However, despite this nominal result on the part of a minority of experimental group participants, the focus group interview showed that, participants' perceptions of the outcomes of the close reading and close writing workshop aligned with that group's increase in SCNA. Alterations noted included improvement in motivation to read; enhanced attention, analysis, comprehension and reflection capacities. These outcomes consolidate the notion that training in close reading and writing actively promotes functional literacy and enhances the development of personal practical analysis and expression strategies.
From the perspective of the facilitator of the workshop, awareness of the distinctions (and similarities) between the use and expectations of close reading in functional literacy, literary criticism and educational psychology helped clarify the orientation of the training. This made it easier to concentrate on the development of personal practical analysis and expression strategies.
Thus, this very short intervention (just three fifty-five-minute workshop sessions introducing a group of university students to close reading and close writing of literary narratives) speaks to the general value of this practice in increasing the quality of the analysis of such narratives. Moreover, the fact that improvements in this capacity were maintained in follow-up speaks to the long-lasting potential of this type of activity. This study could therefore be seen to contribute to the empirical grounding of this approach by showing that introducing university students to close reading and writing of literary narratives (eventually in the context of courses or workshops), seems to offer a means to increase the quality of analysis of narratives, and to retain this improvement over time.
All the same, since that increase did not occur in all participants it seems that this specific intervention is not sufficient for everyone, and would need to be extended or adapted for other characteristics. Moreover, the results of this study must be carefully read considering the reduced number of cases that participated in it. The outcomes may allow generalization to theory but they do not permit a generalization to the population. A future implementation of similar studies with broader samples of participants is required to test this generalization. The use of other and mixed art domains and other narrative analysis techniques would broaden the approach to consider multiple intelligences.

Annex 1: Examples of comments to the excerpt from Harriet Scott Chessman's novel Lydia Cassatt Reading the Morning Paper (2001)
illnesses and deaths-that had occurred in her family. In the end, May warns her sister that they would have to be careful about her diet and they would have to make sure that she rests." 5. Abstract comment (similar to relational comment, but containing elements that can generalise to other contents not present in the text): "Lydia learned last summer that she had Bright's disease-a kidney disease. It was May, apparently her sister, who broke the news to her. For Lydia, everything sounded "unreal" to her ('the whole world around me had disappeared and I felt that [...] I was only made of air'). To make matters worse, Lydia could not reveal what she was thinking or feeling, because her family did not deal with situations of illness and/or death well because of past experiences. From the text it is possible to perceive, for example, how sometimes certain traumas that occur in the core of the family can affect communication between the different members of the family."