Last changed 13 Feb 2012 ............... Length about 6,000 words (40,000 bytes).
(Document started on 23 April 2011.) This is a WWW document maintained by Steve Draper, installed at http://www.psy.gla.ac.uk/~steve/rap/hanscomb.html. You may copy it. How to refer to it.

Web site logical path: [www.psy.gla.ac.uk] [~steve] [rap] [this page]

Hanscomb's virtues: Students' approaches to assignments

By Steve Draper,   Department of Psychology,   University of Glasgow.

Preface

Passport photo This page holds first an essay by me; followed by what I take to be an independently designed intervention addressing the same issue, but with a contrasting attitude Feedback on work related attributes.

The first part of this page is an essay prompted by a talk which I heard on 21 April 2011 by Stuart Hanscomb. It develops my views on the issue, which I think is a new one w.r.t. the educational literature. This essay is far from polished, but I feel I've done enough to sort out my ideas, and to make it more or less readable; so I want to hang it out for a bit to settle. We can hope that this page will be superceded by a paper by Stuart, but I wanted to air the issue at once. All complaints about this page should be directed to me; praise for originality etc. should be directed in the first instance to Stuart.

Stuart gave a talk on this. He also has a draft paper on this.

If you want a summary / abstract of this essay, then jump to the last section.

The underlying phenomemon

Most new ideas are based on some phenomenon, some observation. Educational researchers have most access to their own experience as teachers, and some access to what they hear students mentioning about theirs. What I take to be the underlying heart of this is that, when teachers mark a piece of student work, we often involuntarily perceive or attribute characteristics to the student to do with the way they did it: procrastination, showing contempt for the marker, showing undue conscientiousness, .... This is no surprise to a psychologist familiar with the area of attribution; and the term "fundamental attribution error" names a widespread tendency to attribute our own behaviour to force of external circumstances but others' behaviour to their inherent character.

The educational issue is: should (HE) teachers pay any attention to this? On the one hand, if we want to focus on students' learning we should pay attention to results not to their personal habits such as the clothes they wear, whether they work at night or day, whether they are tall, or bald. On the other hand, employers explicitly ask us to comment in letters of reference on some of these attributes (e.g. diligence, self-starting, timeliness, sickness absences, ...); and most programmes in fact give directive study skill advice. If we were to give feedback based on these attributions, it would be a new dimension to feedback on student work, but a logical continuation of the role of study advisor.

Such attributions and conversations are more likely to arise in cases of personal tuition e.g. PhD supervision, one to one essay feedback in Access courses. But I am periodically aware of making such attributions even when doing anonymous marking of a big pile of exam scripts.

??Note too that these attributions seem often to be inter-personal: that teachers interpret the mode of student work as an overt communication of an attitude of the student towards the teacher (or the work): contempt, thoughtless compliance, joking, .... We take it personally.

On the other hand: "Always remember that it's just as hard to write a bad book as it is to write a good book." [Ken Kesey, "Remember This: Write what you don't know", NYTBR, New York Times Book Reviews 31 Dec 1989]   I take this as a reminder that the most common attributions people (including other students) make about students' writing are wrong.

This is a new, overlooked aspect of feedback

A simple minded view of feedback and learning is that it is all about the learner, not the teacher. It is done to the learner, in order to improve learning outcomes, in a one-way process: telling them what they did wrong so that they will correct their learning. This is an instructivist view, which is bad for learning quality because it ignores the dialogic nature of learning and teaching (which is the heart of Laurillard's model), and is also bad for managing costs because it ignores the nature of the task for teachers.

A number of little-noticed insights into feedback have emerged that illustrate this, and I see Stuart's point as adding to this set. They are characterised by being based on observations about the teacher, not the learner. (They thereby illustrate how unbalanced it is to be "learner-centered": and so ignore the crucial other half of the learning and teaching process.)

  1. Taras (2005) points out that the formative /summative distinction is about effects on the learner, but is nonsense when you look at the teacher. For a teacher, the main time spent is in reading the student's work, and then a judgement (summative) comes into their head. Witholding the judgement simply means witholding information from the learner; adding formative comments is relatively little extra work. (Or at least, not doing so wastes the time spent reading and making the judgement; even if you are a tutor who writes an essay in reply to an essay.) Summative and formative are not separate from the viewpoint of teachers' work.

  2. Peter Elbow points out that we have not one but two critical modes: in one (judgemental, authoritative, "constructive") we tell the learner what they should have done; in the other, we tell them what we felt when we read their work (describing our personal feelings and interpretations). The latter "Reader-Based Feedback" mode is forced on us when we don't know what the author intended to communicate; but it is also much less affronting for authors sensitive to criticism. Frequently published advice or even directions about writing feedback for students, that requires it to be "constructive", presupposes that all students write so well that you know what they meant at all times, and also assumes that all students are infinitely resilient (i.e. insensitive) to criticism. [The same literature repeatedly expresses bafflement or outrage that students frequently do not attend to or even read "formative" feedback, as if they didn't take it personally; yet also expects students to have an instrinsic interest in learning (i.e. to identify with it personally).]

    A more sensible post-Elbow view is that we have these two modes in our arsenal, and need to select the mode in response both to the particular work, and the particular student.

    The Reader-Based Feedback voice can express feedback successfully even to very sensitive recipients. In normal life, I find myself naturally using a mixture of the two voices when reviewing for journals; but feedback to students is often solely in the authoritative voice, while the aspects of the students' work actually range from the factual (which are right or wrong) to stylistic matters which are not agreed on within the discipline and so for which an authoritative voice is in fact bogus.

  3. Work by Beryl Plimmer drew my attention to how markers are writing to multiple audiences. When I mark an essay I am generally a) writing private notes on new ideas (or doubts) the student has given me to follow up for my own benefit later; and for making adjustments to the course next time. b) writing private notes on my view of this piece of work: it will be in private jargon and abbreviations, and uncensored as to content. It will be used later for discussions with colleagues on mark adjustments, and as a basis for writing feedback to the student (I have to work out what I think first, before I can explain some subset of that to the student). c) writing to the student: formative feedback. d) writing at least a mark to the department bureaucracy e.g. a marks database. e) writing comments on the script itself e.g. spelling errors, or the location to which some comments apply.

    In addition to these five or so simultaneous writing tasks to five different audiences on five different bits of paper or software windows, there may be multiple simultaneous reading tasks. In Plimmer's case of marking computer programming assignments there were source code, test suite output reports, design documentation. In the case of essay-like assignments, there is usually at least the essay body and the bibliography to turn between repeatedly.

    As soon as Plimmer provided software that reduced the labour of managing all this, the quality of feedback to students, as judged by the students, went up markedly; thus illustrating the importance of analysing and supporting the teacher side of feedback provision.

  4. Hanscomb points out that markers automatically form these judgements / attributions / perceptions of students from their written work. We should pay attention to this, just as we should pay attention to Elbow's alternative voices, and Plimmer's multiple outputs of marking.

I'm thus sympathetic to Stuart's insight because it seems to add to this collection of observations about what teachers really do when they respond to students' work (as opposed to what policy or theory says feedback should be about). If we, as teachers, have these perceptions of students, make these attributions, then what should we do with them? Are they, or could we make them, valuable?

Stuart's list

Stuart gives examples of these percepts or attributions, which he calls "virtues".

His list of candidate virtues (my names; he had 2 line descriptions):

Other examples

Other examples, at least of where such perceptions are not only made but voiced, tend to come in one-to-one project supervision. A tutor might say "you seem very involved in this topic/ aspect: do you think overly so?"; or "you've completely skipped over that aspect yet I've heard you speak passionately about it: why?".

Blay Whitby suggests something like: A virtue corresponding to the basic scholarly requirement of basing your (the student's) work in the academic community: of showing awareness (e.g. in a literature review) of how your ideas fit into a landscape of existing ideas. Seen as a centre-optimal virtue à la Aristotle, then the two (bad) extremes would be a) polemics that ignore all other views and contributions, and b) exhaustive, pedantic, and undiscriminating bibliographies.

What kind of things might these be?

He calls these attributions "virtues". However they, or some of them, seem equally plausibly to be other kinds of traits.

The nature of these types

At one point Stuart contrasts virtues to skills. This is right, but the important general point is that you cannot have a graduate attribute, a virtue, an approach, a strength, nor a personality trait without at least one corresponding skill to exhibit it. You more often have a set of such skills for each attribute, but without at least one you can't manifest it. For some virtues, e.g. courage, this isn't a barrier: you can see even in an infant whether their behaviour at each moment is high or low in courage. But for other attributes, say flexibility w.r.t. norms, it is not obvious to everyone how to manifest these.
  1. First you must learn at least one skill or method per attribute.
  2. Second if you have more than one applicable method, there is the question of how you choose which one to apply. One constraint (b1) is what set of methods an individual has. Another (b2) is the context: the demands of the task, and the limitations of opportunity. A third influence (b3) is the preferences of the individual, if and when the choice isn't already decided by what is possible and required.
  3. Thirdly is the issue of what kind of pattern, if any, is revealed by such preferences.

Because of (a) they aren't fixed traits but progressively developing abilities or methods. Because of (b) -- choice of methods -- they aren't really attributes of the agent (person), but "relations" i.e. determined by the combination of agent and context, not by either one alone.

The nature of these types (2)

All the above types are habits of character; all require skills in order to be manifested; and so all require learning ("cultivation") which strongly modifies when they can be acted on. They differ in whether:

  • The normal sense of the English word "virtue" is unipolar (having it is good, not having it is bad).
  • Aristotle's virtues have the centre as good (valued) and the two (bipolar) extremes are less so e.g. cowardice - courage - recklessness (or greed - healthy appetite - eating disorder).
  • Hanscomb's virtues however are unipolar: he wants his students to show responsibility, flexibility, resilience, etc.
  • The VIA strengths are seen like this (unipolar, valued), but with the attitude that there are individual differences, and the focus is on which strength an individual has (excels in).
  • Personality ("big five") traits are bipolar like Aristotle's virtues, though mostly not valued but just differences.

    My conclusion on the nature of these types

    I think Hanscomb's virtues are "approaches"; "approaches to assignments" (as opposed to "approaches to learning" as in the deep and surface learning literature).

    "Approach" is the word used for the alternative methods or strategies learners show towards learning, called deep and surface learning. It is tempting to see these as traits and talk of "shallow learners", but this is wrong. Learners typically can do both, and select one depending on the context. Teachers can usually force almost all learners into shallow learning, though not the other way round. Thus approaches are not attributes of the learner, but a relationship determined by the combination of learner and context. (Biggs (2001) puts this best.)

    A strategy or approach is a high level planning decision, and often involves deciding on priorities, the relative importance of various parameters. For instance in project management, should quality, quantity or cost be optimised? In war, what are the war aims? Should territory, civilian deaths, or keeping the army intact be most important?

    In learning, the deep-surface contrast seems to be about whether understanding is the aim, or learning as measured by a specific form of test. In cases where a course has multiple forms of assessment (e.g. weekly quiz, end of semester exam, progression to a higher degree), a somewhat different kind of strategic choice must be made by students as discussed by Snyder (1971). In Hanscomb's cases it is about getting an assignment done: so the form of test is fixed, the deadline is fixed, and the strategic choices might be about time, cost, quality. However only two of his virtues seem to be of this kind: flexibility, and sloppiness. Three others seem to be about how the student responds to, not one assignment, but the cycle of repeated assignments and feedback (perseverence, resilience, defensiveness). The other three seem more to do with links between personality traits and Perry-like features of students' "epistemology": responsibility, open mindedness, respect for others' views.

    However they are all to do with, not regulating learning (as in how a student does, and regulates, their revision), but with doing an assessed assignment. Thus I view these as approaches, not traits; and as distinct from approaches to learning (deep and surface). This is the second lesson I learned from Stuart. It is a new area of student choice.

    Perry related

    Some of Hanscomb's virtues seem related to the Perry issue, but they are simultaneously to do with the demands of essays in the social sciences, and with the interpersonal attitudes that students should have towards peers and teachers. Open-mindedness is about whether you can learn from arguments and what others say, or are stuck with the judgement you first thought of.

    Put another way: There is a deep issue in education associated with Perry's (1968) work about how a learner relates to differences in views, opinions, but it has four different instantiations.

    1. The nature of the material to be learned. Perry focussed on this at the level of subject matter content: where discipline areas require their learners to understand and write about differences in views of experts in the field. However he wrongly assumed that a person has a single consistent epistemological position about the nature of knowledge (e.g. relativist), whereas it is nearer the truth to say that one of the things we quickly learn about each topic is its status. For instance almost everyone sees music as something where views differ but there is little prospect of arguing someone into or out of likeing one style of music. In contrast, almost everyone sees tide tables as a matter of fact: neither arguing about it, nor any technology available, makes any difference to when in fact the tide will come in.
    2. Critical thinking. Closely related to this is to see it as "critical thinking": where the required style of discourse is to discuss alternative views and the differences in quality of evidence for each. This is in effect a skill or procedure many students must learn, like calculus for scientists, regardless of their preferences.
    3. Peer interaction styles. But it is also important in how a student interacts with their peers in discussion: a matter in part of social conduct. Newman saw this as the central value of universities: bringing students in contact with people of different opinions but similar status, intelligence, and well-informedness. This is partly a training for life, where you have to deal with such differences in opinion; and where justifying your opinion is often required. But it is also an opportunity to learn. And a matter of social conduct: showing respect for others' views is a requirement in many contexts.
    4. Personal identity. But it also has a personal aspect. When your views are attacked, do you feel that it is your identity that is attacked, undermined? (On the other hand, if you are detached from the views you discuss, does that mean you appear insincere socially, internally feel empty and unmotivated about studying the area, and see arguments as being merely a game like chess, unrelated to practical action?)

    These four aspects are by no means necessarily unified in each person. We may express our perceptions of a student in terms of personal identity or attributes, yet want to teach them: i.e. want to change not their identity but their behaviour.

    If we take this view, then what should we do? I'm not aware of good literature on active interventions to promote progression to better Perry levels. But presumably the intervention would be (partly) to interpret student work in terms of the Perry related stage or attitudes it shows, and to do active tutorial interventions to try to progress the student in this respect.

    We should give systematic feedback on study skills, self-regulation

    If we accept that we give any teaching at all about study skills, methods, and personal development, then we shouldn't just give exposition but feedback. Feedback has to be in response to some kind of student work; and Hanscomb points out that teachers automatically form the basic perceptions. How might we organise this?

    Pre-facing work with questions requesting feedback. This is the technique of elective feedback. We could extend this by requiring students to submit, along with their essay, an account (learning diary?) of what work they did for it, and how they organised that effort. A time management report: but probably the quantitative report on time is less important than the qualitative report of how the work was broken down into parts, and what those parts were. Then the tutor could give feedback on this management (study skill) aspect of the work.

    The Patchwork Text technique could be relevant here: requiring learners to make entries each week into their learning diary; then review them at half term.

    This kind of feedback is about attending less to the current product by a learner (how to re-edit this particular essay); and more to general approaches in the learner that led to it, and seeing how to modify these for the future.

    Is this intrusive? or is it taking an interest in what students do? Draper (2008) coined the phrase "Teacher monitoring" to discuss the indications that students may welcome this attention even when it is not kind or "supportive" in emotional flavour.

    Conclusion

    So: Stuart suggests giving explicit feedback on three neglected areas; and to do so by expressing it in terms of character attributions. This could be better than in technical terms. It could be done, following Elbow, not in the voice of expertise, but in the voice of expressing personal impressions; thus largely obviating the danger of sounding arrogant, impertinent, ....

    At interesting suggestion is that the process of starting from these instantaneous and unconsidered judgemental perceptions and considering them seriously, is that a marker might learn as much about themselves as about the students. This might well include a process leading the tutor to become more articulate about the implicit intellectual values they hold and would wish their students to internalise.

    Summary

    References

  • Biggs,J. (2001) "Enhancing learning: A matter of style or approach" ch.4 pp.73-102 in R.J.Sternberg & Li-Fang Zhang (2001) (eds.) Perspectives on thinking, learning, and cognitive styles (London: LEA) [Argument that deep& shallow learning is not a trait]

  • Draper,S.W. (2008) "Learning and Community" pp.132-144 in A.Comrie, N.J.Mayes, J.T.Mayes & K.Smyth (Editors) Learners in the Co-Creation of Knowledge: Proceedings of the LICK 2008 Symposium, Edinburgh 30 October 2008 (Edinburgh: Napier University). [Teacher monitoring]

  • Peter Elbow:   wikiP   Partial bibliography

  • Stuart Hanscomb

  • Newman,J.H. (1852/1976) [ed. I.T.Ker] The idea of a university (Oxford: Oxford University Press).

  • Perry, W.G. (1968/70) Forms of intellectual and ethical development in the college years (New York: Holt, Rhinehart and Winston)

  • Connected vs. separate learning. See:

  • Plimmer, Beryl references: http://www.cs.auckland.ac.nz/research/hci/digital_ink/annotation_tools/pen_marked.shtml

  • Snyder,Benson R. (1971) The hidden curriculum (MIT press; Boston, Mass.)

  • Strengths

  • A tone of voice (cf. Vygotsky)? An attitude to an activity, that could quite well be very different, but has large consequences when adopted.

  • Taras,M. (2005) "Assessment -- summative and formative -- some theoretical reflections" British Journal of Educational Studies vol.53 no.4, pp.466-478

    Vyg? or neo-Vyg ref?

    What original thoughts do I think I have here?

    1. New 4-level Perry classification
    2. The nature of traits vs.approaches / habits of character
    3. A classification of the forms these can be expressed in
    4. (Recognising Stuart's identification of a new phenomenon.)
    5. Articulating what kind of thing this is; listing others of the same family.

    ?More partial associations

    Belenky connected knowing.

    Feedback on work related attributes

    Passport photo Lorna Morrow is developing a feedback sheet and procedure for doing a version of Hanscomb's idea, but focussed on work related attributes, in this department. The spirit of her approach is quite different. No rumination on the nature of virtues or traits, but direct telling students of a tutor's judgement, phrased as authoritative truth, just like an essay mark.

    This brings out how this proposed learning & teaching process is related to:

    A possible process

    After a brainstorm (ambiguity intended), here is an elaborated possible procedure based on this idea.
    1. Student is required to come up with (say) 7 headings of their own; and their own self-evaluation against those headings.
    2. Tutor comments on these. Then comes up with additional headings (perhaps from a standard list), and the tutor's comments on the student under each of these.

    3. A departmental list based on the headings which the department staff assembled based on their recent collective experience of headings in requests for references for students. Or Lorna's list which is close to this. Tutor and/or student translate or carry over the comments from earlier steps to fill in these headings.
    4. Graduate attributes (the university's set). Tutor and/or student translate or carry over the comments from earlier steps to fill in these headings.

    Process v.2: Linking behaviour and attribute

    Another issue to consider is the advantage of linking each attribute to an observed behaviour, not just to a perceived degree as in the appraisal-style process above: e.g. "quite conscientious". An example of this version would be "emailing back revisions in response to every bit of feedback given shows that you are conscientious". (For more examples, if you don't mind a dubious reference, see the section on "positive programming" here and adapt the examples for our purposes here.)

    Process v.3: Use Elbow's Reader-Based Feedback voice

    The processes sketched above use an absolutist voice, phrasing things as if the tutor's attributions of the student's traits were unquestionably veridicial. An alternative here, as in fact in any feedback, is to use the "other" voice of feedback, described by Peter Elbow (see above), and which he called "Reader-Based Feedback". It phrases nothing as an attribute, presupposed true, of the student; and everything as a feeling of the reader/tutor e.g. "your emailing back revisions in response to every bit of feedback given makes me think you are conscientious"; or "I was excited by your introduction, but felt lost by the end of section 2".

    Both because tutor attributions of student traits are going to have low reliability, and because they are personal and likely to be emotionally laden for the recipient, the Reader-Based Feedback voice is probably much more suitable for feedback on work related attributes than the authoritative voice.

    A quite different process

    Alternatively, suppose we want to give reliable feedback on traits. Then the view would be to reject all the above suggestions because when we provide feedback from 'thin slices' of behaviour interpreted from 'signs' in behavioural output on traits/skills that have alternative, more reliable and valid measures available, we do ourselves and our students a disservice.

    The implication is that we should actually measure the traits of interest using validated procedures from psychology. We could then both give real personal feedback (not vent our amateur impressions), and give more evidence-based references for future employers.

    Web site logical path: [www.psy.gla.ac.uk] [~steve] [rap] [this page]
    [Top of this page]