Common misunderstandings of phenomenographic research in higher education

ABSTRACT

This paper outlines the impact of phenomenography on higher education research and academic development. Interest in phenomenography as an educational research methodology continues to grow, but with interest growing faster than the number of experienced researchers, some misunderstandings of the approach have arisen and been circulating in discussions and publications, becoming self-reinforcing. This can cause misinterpretations of the meaning and implications of phenomenographic research findings, plus potential confusion for research students and others new to the research approach. This paper describes eight prevalent misunderstandings of phenomenography, explains what is missing in those ways of thinking about phenomenographic research, and speculates on the sources of such misunderstandings. A unique aspect of this paper is the focus on what phenomenograhy is not, as well as what it is, which is congruent with a phenomenographic approach to learning and development.

KEYWORDS:

• phenomenography
• misunderstandings
• higher education research
• research methodology

Why is phenomenography significant in higher education research?

Phenomenography is unique within higher education research as the only methodology that has developed within the context of higher education itself, rather than being imported from another field and then applied to higher education (Tight, 2016). In addition, in Australia and the UK in particular, it has been highly influential amongst academic developers in higher education institutions and in the design of Graduate Certificates of Higher Education and other programmes of educational development for academic staff (Kandlbinder & Peseta, 2011a; Webb, 1997).

Phenomenography has had a major impact on the development of the field of higher education research. In a 2007 review of 10 years of publications in the HERD journal, Kandlbinder (2011) identified the 10 most cited authors in the history of the journal, and found that the first seven were all associated with phenomenographic research.¹ Furthermore, in a historical anthology designed to capture the conceptual development of the field of higher education research over time, Kandlbinder and Peseta (2011a) used bibliographic review and citation analysis to identify 15 particularly high impact articles published in HERD over the preceding three decades. The articles were grouped into five themes, representing key concepts taught in Graduate Certificates, as identified from a survey of Coordinators of Graduate Certificates in Australasia and the UK.

Within the anthology, phenomenographic research was represented in articles from three of the five themes, with all three described as having ‘become central to the field’ of higher education teaching and learning (Kandlbinder & Peseta, 2011b, p. 9):

• student approaches to learning;

• constructive alignment; and

• scholarship of teaching.

This further illustrates the deep influence of phenomenography on the development of higher education as a field of research and practice.

Impact on educational research

Phenomenography is perhaps best known for establishing the now famous distinction between deep and surface approaches to learning (Marton & Säljö, 1976), which has had a major impact on higher education research and pedagogical practice, especially in Australia, New Zealand and the UK (Marton et al., [1984] 1997; Ramsden, [1992] 2003; Prosser & Trigwell, 1999; Biggs & Tang [1999, 2003] 2007; Kandlbinder & Peseta, 2011a).

As described by Biggs and Tang (1999), ‘this series of studies [by Marton and colleagues] struck a chord with ongoing work in other countries; in particular with that of Entwistle in the UK and Biggs in Australia’ (p. 61). Both Biggs and Entwistle were working independently on the development of inventories to measure approaches to learning, but from different conceptual frameworks to phenomenography – individual differences psychology (Entwistle) and cognitive psychology (Biggs). The study processes or orientations they were investigating were conceived of from a psychological perspective, as arising from relatively stable psychological processes or constructs. This is in contrast to the relational perspective espoused by phenomenography, where approaches to learning were conceived of as part of a set of logical relationships between students’ intentions when learning, the approach they take to their learning, and the resulting outcomes of their learning. Nevertheless, the results of the three strands of research showed a substantial degree of alignment.

But even more important than the deep-surface distinction itself was that Marton’s phenomenographic approach to researching student learning helped to inspire a paradigm shift in the foci of higher education research, from the then dominant approach of experimental investigations of the quantity of learning, undertaken in controlled settings, to naturalistic investigations of the quality of learning, undertaken in natural educational settings (Ramsden, 1985; Entwistle, 1991; Entwistle, [1984] 1997; Svennson, 1997).²

The way in which the research was carried out, with its focus on understanding the experience of learning from the students’ perspective, promoted ‘a different way of thinking about teaching and learning’ (Ramsden, 1985, p. 52) that was broadly influential, even amongst those not engaged in phenomenographic research itself. For example, Richardson (1999) described phenomenographic research as having ‘revolutionized the way in which both researchers and teachers think about the process and the outcome of learning in higher education’ (p. 72) – even though he was critiquing phenomenography at the time. And Biggs acknowledges the influence of phenomenography on the development of his highly influential notion of ‘constructive alignment’ in university teaching and learning (Biggs & Tang, 1999; Biggs & Tang, [1999, 2003] 2007).

Impact on academic development programmes

Phenomenography was also influential in the design of academic development activities and graduate programmes on higher education teaching and learning for academic staff. This was due not just to the perceived value of student learning research to academic development, but to the findings of phenomenographic research on conceptions of learning, teaching and teacher development, as well as students’ understanding of different disciplinary concepts.

With regard to learning, an empirical and logical relationship was established over a series of studies between students’ conceptions of learning, approach to learning and understanding of disciplinary concepts (Marton et al., [1984] 1997; Prosser & Trigwell, 1999). With regard to teaching and learning, an empirical and logical relationship was established over a series of studies between teachers’ conceptions of teaching, approach to teaching development and their students’ conceptions of learning (Åkerlind, 2003, 2004, 2007; Prosser & Trigwell, 1999; Trigwell & Prosser, 2020).

This set of interrelated findings led a number of researchers and academic developers to argue that the best approach to take towards education and development programmes for university teachers is to focus on developing their conceptual understanding of the nature of teaching and learning, in contrast to the traditional approach of focusing on teaching methods (Åkerlind, 2003, 2004, 2008; Booth & Anderberg, 2005; Kember, 1997; Martin & Ramsden, 1992; Prosser & Trigwell, 1997, 1999; Trigwell & Prosser, 1996, 2020).

Ongoing growth of phenomenographic research

Whilst the passage of time has reduced the ‘dominance’ (Webb, 1997) of phenomenography in academic development, it continues to be an influential research approach with which most academic developers (at least in Australasia, UK and Europe) will be familiar. In addition, phenomenography’s use in educational research has continued to rise internationally. In a review of the development of phenomenography as an educational research method, Tight (2016) records an ongoing rise in the number of publications that refer to or use phenomenography across the decades.

Sadly, this rise in popularity is also being accompanied by a spread of misunderstandings of several aspects of phenomenography. Most concerningly, such misunderstandings are even being seen in methodological publications trying to describe the research approach for the benefit of others (eg., Feldon & Tofel-Grehl, 2018; Hajar, 2021; Straub, 2021). I have previously argued that this spread of misunderstandings is due to three main factors (Åkerlind, 2022):

1. that many try to use phenomenographic methods based solely on reading about the approach, without the assistance of an experienced colleague, mentor or supervisor;

2. that descriptions and expectations of the approach have changed over time, but many researchers still use the early and least developed descriptions of the approach as their main reference;

3. that the common practice of self-designation of areas of expertise by reviewers leads to failures in the journal review process such that methodological errors in phenomenographic research are not always caught during review.

The ongoing spread of misunderstandings of phenomenography makes further explanations of phenomenography by experts in the field imperative. That is the aim of this paper. However, I am going to do so in an unusual way, spending as much time describing what phenomenography is not, as what it is, by concentrating on common misunderstandings of the research approach.

This focus on highlighting misunderstandings in describing phenomenography makes use of the notion of ‘contrast’, as described in the ‘Variation Theory of learning’ derived from phenomenography (Marton, 2015). The assumption of Variation Theory is that any phenomenon is better understood in contrast to other phenomena than when described in isolation. This is because contrast helps to draw attention to certain aspects of the phenomenon that may be less apparent without something to contrast it with. In other words, contrast helps us to discern aspects of a phenomenon that we otherwise may not notice. So, describing phenomenography in terms of what it is not highlights different aspects of phenomenography to describing it in terms of what it is.

Common misunderstandings of phenomenography

Below I present a number of common misunderstandings of phenomenography. In coming up with this list, I draw on my experience from reviewing journal submissions, examining doctoral theses and participating in conference discussions more than my experience of the published literature, where such misunderstandings have often been filtered by the journal review process. Nevertheless, I illustrate each misunderstanding with two or three publications, showing that the review process is only partially effective. I also focus only on misunderstandings that I have seen multiple times, not just once or twice.

Misunderstanding 1: that ‘approaches to learning’ research constitutes phenomenographic research

It is ironic that phenomenography initially became well known due to the popularity of research on deep and surface approaches to learning, as a generalisable framework for viewing approaches to learning across many contexts. The irony is that this decontextualization of the deep-surface distinction from the specific content of learning removes much of its phenomenographic nature. In fact, Marton specifically suggested that approaches to learning would be expected to take on different characteristics in different learning contexts, and consequently advocated for studies that investigated approaches to learning separately in different disciplinary areas (as illustrated in Marshall & Case, 2005).

In fact, very little of the approaches to learning literature is phenomenographic in nature, having moved largely in the direction of correlational studies of the impact of learning context on approaches to learning, and psychological constructs associated with approaches to learning (e.g., Entwistle, 1991; Ramsden, [1992] 2003; Prosser & Trigwell, 1999; Marshall & Case, 2005; Trigwell & Prosser, 2020). Nevertheless, the association between phenomenography and the demonstration of deep and surface approaches to learning has led some researchers to the misunderstanding that approaches to learning research represents phenomenographic research (e.g., Barattucci & Bocciolesi, 2018; Vermunt, 1996; Webb, 1997).

In the study by Marton and Säljö (1976) that first distinguished between deep and surface approaches to learning, they reported two main findings:

1. that university students who read the same text arrived at qualitatively different understandings of the content;

2. that the differences in understanding were logically aligned with differences in the way students set about learning, i.e., that there is a relationship between learning process and learning outcome.

The different ways in which students set about learning came to be called ‘approaches to learning’, and this second finding from the study went on to become the basis for future research on deep and surface approaches to learning, much of which was not phenomenographic in nature. The phenomenographic focus was much more on the dialectical nature of the relationship between approaches to learning (the act of learning) and learning outcomes (the content of learning), rather than on the specific nature of the approaches to learning found. The ‘act’ and ‘content’ of learning were subsequently positioned as two parts of the same learning whole, i.e., the ‘how’ and ‘what’ of learning (Marton & Säljö, [1984] 1997; Marton, 1994; Marton & Booth, 1997).

Meanwhile, the first finding from the study went on to form the basis of most phenomenographic research – that is, that there will be a limited number of qualitatively different ways of understanding any phenomenon in the world, and that empirically investigating different understandings of disciplinary concepts and principles would ‘undoubtedly prove fruitful for teaching’ (Marton & Säljö, 1976, p. 10).

Misunderstanding 2: that the aim of phenomenography is solely to investigate different ways of understanding a phenomenon

Another common misunderstanding of phenomenography is that it is simply an investigation of variation in understandings of the same phenomena, without realising that explication of structural relationships (typically a hierarchical structure) connecting the different understandings is also essential. This misunderstanding is most evident in empirical studies that claim to have used phenomenographic methods, but report only variation in understandings in their findings, without going on to establish a structure or relationships linking the different understandings found. I often see this when reviewing journal submissions and examining doctoral theses, but sadly, it is also quite prevalent in the published literature (e.g., Feldon, et al., 2015; Hajar, 2020; Hubinette et al., 2014; Schaub-de Jong et al., 2009; Vermunt, 1996).

The likely cause of this misunderstanding is that the object of research in phenomenography is most commonly presented as ‘descriptions of the qualitatively different ways in which people perceive and understand their reality’ (Marton, 1981, p. 177) or ‘mapping the qualitatively different ways in which people experience … phenomena in the world around them’ (Marton, 1986, p. 31). Unfortunately, such descriptions highlight the phenomenographic search for variation, but not the simultaneous search for structure. It is because of the focus on producing a structured set of outcomes that phenomenography uses terms like ‘mapping’ and refers to an ‘outcome space’ rather than just outcomes.

Marton’s elaborations later in both papers highlight that clarification of structural relationships between understandings is also an expected part of the research approach, for example,

The results of phenomenographic research are the categorizations of descriptions [of the phenomenon] … each category is a potential part of a larger structure in which the category is related to other categories of descriptions. It is a goal of phenomenography to discover the structural framework within which various categories of descriptions exist. (Marton, 1986, pp. 33–34)

However, such elaborations are easy to miss when reading these early articles, because they are mentioned only briefly. The importance of structure becomes more obvious in later works, where it is described in more detail. Meanwhile, because Marton (1981) and (1986) were the first widely published papers on phenomenography, they are also the two most cited papers in phenomenographic publications, unintentionally encouraging this misunderstanding of phenomenography.

Misunderstanding 3: that the hierarchical ordering of different understandings in phenomenographic outcomes is based simply on a value judgment

Even where researchers do recognise that phenomenographic outcomes are expected to highlight structural relationships between different ways of understanding the same phenomenon, the nature of the relationships and the typically hierarchical structure that emerges in a phenomenographic outcome space is often misunderstood. The hierarchy is often mistakenly seen as a simple ordering of the understandings, based on authorised understandings or value judgments (e.g., Ashworth & Lucas, 1998; Kember, 1997; Webb, 1997). In contrast, the structural relationship between understandings in a hierarchically structured outcome space is actually one of inclusivity, in the sense that understandings higher in the hierarchy have been placed there because they include awareness of understandings lower in the hierarchy, but not vice versa.

Because they are inclusive, understandings higher in the hierarchy are inevitably more complex than those lower in the hierarchy, and this increased complexity is seen as enabling more flexible and powerful ways of operating in the world (Marton, 1994; Marton & Booth, 1997). So, in this sense, the most inclusive understandings may also be regarded as the ‘best’, or most sophisticated, way of understanding the phenomenon, but the ordering is based on inclusivity, not simply on a value judgment.

Misunderstanding 4: that the interpretive role of the phenomenographic researcher is a subjective one

Phenomenography assumes a non-dualistic epistemology, in which individuals’ understanding of a phenomenon is seen as a relationship between the individual and the phenomenon being experienced:

There is not a real world ‘out there’ and a subjective world ‘in here’. The world [as experienced] is not constructed by the learner, nor is it imposed upon her; it is constituted as an internal relation between them. (Marton & Booth, 1997, p. 13)

If this applies to all phenomena, then this inevitably includes research phenomena. In other words, researchers’ experience of research data will inevitably be a relationship between the researcher and the data. So, in phenomenography, research interpretations are regarded not as subjective or objective, but as relational: ‘The experiences and understandings are jointly constituted by interviewer and interviewee’ (Marton, 1994, p. 4427).

In contrast, there is a common misunderstanding of phenomenographic research interpretations as being simply subjective. This can be seen most obviously in some critiques of phenomenography (e.g., Webb, 1997). But, more subtly, it can also be seen in guidelines on conducting phenomenographic research that describe the importance of avoiding subjectivity through ‘bracketing’, but without simultaneously acknowledging the relational nature of phenomenographic analysis (e.g., Ashworth & Lucas, 1998, 2000; Hajar, 2021). This, then, has implications for what are regarded as appropriate approaches to maintaining rigour in phenomenographic research, in particular the pros and cons of ‘bracketing’ (which focuses on avoiding subjectivity) vs ‘interpretive awareness’ (which focuses on acknowledging subjectivity) – as I have discussed elsewhere (Åkerlind, [2005] 2012, 2022).

Misunderstanding 5: that the understandings expressed by interviewees in a phenomenographic interview can be taken at face value

It is not uncommon to see descriptions of the phenomenographic interview that state the need to bring interviewees into a ‘state of meta-awareness’ with respect to the phenomenon under study. This claim comes from Marton and Booth (1997), though in other work, Marton simply refers to the importance of encouraging participants to reflect on their experience as actively as possible (e.g., Marton, 1994). Although Marton uses ‘reflection’ and ‘meta-awareness’ interchangeably across his writings, the use of the term, meta-awareness, has turned out to be unfortunate, because it is a more ambiguous term and, if used without explanation, can create the false impression that phenomenographic researchers think interview participants can become so aware of their way of thinking about a phenomenon that they become able to explicitly describe it to the researcher, and all the researcher needs to do is record that description.

But this idea, even if it were possible, goes against the methodological assumptions underlying phenomenography, in particular that what one person says about a phenomenon can only be understood in comparison with what other people say about it (as described above). So, individual interview transcripts can never be interpreted in isolation. During analysis, an individual interview transcript may be ‘allocated’ to a particular category of understanding, in the sense that it is seen as illustrating the way of understanding represented by that category³, but this does not mean that that way of understanding has been explicitly articulated within the transcript. In fact, meaning is much more likely to be expressed implicitly than explicitly in interviews.

This leads me to a related misunderstanding of the phenomenographic interview, that the researcher takes what an interviewee says at face value (Säljö, 1996; Webb, 1997). This is far from the case. Interviewees are not expected to be able to express their understanding explicitly. Nor is what they are able to explicitly say about their understanding taken at face value in an interview. It is well accepted in phenomenography that the same words or phrases can be used to express different meanings, while different words and phrases can be used to express the same meaning.

Consequently, a major role of the interviewer is to continually check for the meaning underlying the language used by the interviewee, through follow-up questions such as: ‘What do you mean by that?’; ‘Can you give me a concrete example of that?’; ‘Why did you do it that way rather than some other way?’; ‘What are you trying to achieve through that?’; ‘Why is that important to you?’; What would happen if you didn’t do that?’; ‘How does what you are saying now relate to what you said earlier?’. Indeed, I often argue that the follow-up probes are more important than the main questions in a phenomenographic interview.

Misunderstanding 6: that phenomenography investigates individual participants’ ways of understanding

Because phenomenographic data collection typically involves asking individuals about their experiences, it is often incorrectly assumed that phenomenography studies individuals’ ways of understanding a phenomenon. In contrast, one of the ways in which phenomenography defines itself is as investigating collective understanding, not individual understanding (Marton, 1981, 1986; Marton & Booth, 1997).

Although the collective view can only be accessed via individual views, phenomenographic data analyses and research outcomes do not emphasise the experience of individual interviewees, but rather the collective experience of the interview group as a whole. This is another reason for the importance placed by phenomenography on searching for structural relationships that link the different ways of understanding that emerge, so the understandings can be looked at as a collective whole.

Phenomenography argues that people experience the world differently because human experience is always partial, that at any one point in time and context, people will discern and experience different aspects (i.e., some aspects and not others) of a phenomenon. This means that:

the limited number of qualitatively different ways in which something is experienced can be understood in terms of which constituent parts or aspects are discerned and appear simultaneously in people’s awareness. (Marton & Booth, 1997, p. 107)

This leads to the assumption of a part-whole structure to human experience or understanding, with different ways of understanding the same concept seen as interconnected parts of the whole composite of understandings, ‘the meaning of one bit derived from the meaning of and lending meaning to the rest’ (Marton & Booth, 1997, p. 124).

Consequently, phenomenography assumes that what is significant in what one person says about a phenomenon can only be understood in comparison with what other people say about it. So, the categories of understanding that result from phenomenographic analysis do not describe a way of understanding a phenomenon in the way an individual would, but in terms of the aspects of the phenomenon that have been highlighted by its similarities and differences to other ways of understanding.

This focus on similarities and differences is what enables the inclusively expanding structure of the outcome space. It also explains the ‘stripped’ or minimised nature of phenomenographic descriptions of ways of understanding, because they represent a way of experiencing reduced to its key critical aspects. Rather than focusing on the endless variation inherent in the richness of individual experience, phenomenographic research focuses on identifying what is critical for distinguishing one way of understanding from a qualitatively different way, by describing the minimum aspects necessary for drawing such distinctions.

Misunderstanding 7: that phenomenography adopts a similar cognitive epistemology to constructivism

Biggs claims that there are two main theories of teaching within the student learning paradigm, phenomenography and constructivism (Biggs & Tang, 1999; Biggs & Tang, [1999, 2003] 2007). What the theories share is the assumption that meaning is not transmitted by direct instruction, but created by students during their learning activities; and that learning is about qualitative conceptual change, not quantitative accumulation of information. From both perspectives, learning is seen as resulting from interactions with the world, that ‘as we learn, our conceptions of phenomena change, and we see the world differently’ (Biggs & Tang, 1999, p. 60).

Where phenomenography and constructivism diverge is in the epistemological assumptions as to the nature of a conception and conceptual change; i.e., what happens when experience leads to a change in conceptions. Marton is explicit in distinguishing phenomenography from constructivist and cognitive models⁴ (Marton, 1994; Marton & Booth, 1997).

What, then, is a conception of something, or a way of experiencing something? … It is not a mental representation or a cognitive structure; it is a way of being aware of something. (Marton, 1994, pp. 4425–4426)

Nevertheless, some authors describe phenomenography in constructivist/cognitive terms (e.g., Straub, 2021; Vermunt, 1996).

As mentioned above, phenomenography adopts a relational epistemology, positing that human knowledge of the world is non-dualistic in nature. In contrast, constructivism takes a dualistic stance, with a focus on the subjective world, in terms of individuals’ internal constructions or representations of an external reality. This may be further contrasted with a positivist epistemology, which seeks to study an external reality, independent of human interpretation of it.

The difference between the epistemological assumptions of phenomenography and constructivism becomes even more apparent when one considers their different implications for the nature of conceptual development (i.e., learning). Constructivist epistemological assumptions lead to taking a conceptual change approach to conceptual development. By contrast, phenomenographic epistemological assumptions lead to taking what I have termed a conceptual expansion approach to conceptual development (Åkerlind, 2008).

A conceptual change approach is based on a view of learning as involving the replacement of one system of beliefs or concepts with another. Conceptual change is seen as occurring through reflection on experience, where new information is incorporated into pre-existing knowledge schemas. The assumption is of a relatively rational set of beliefs and encouraging conceptual change through facilitating individuals’ ability to reject, on some logical basis, existing erroneous conceptions in favour of new, more accurate conceptions. Pedagogical strategies thus focus on active and experiential learning and the use of anomalies, analogies and exemplars in facilitating conceptual change.

This approach may be contrasted with a phenomenographic, or conceptual expansion, approach to conceptual development. From a phenomenographic perspective, less sophisticated conceptions are regarded not so much as wrong, but as incomplete or limited, lacking awareness of key aspects of the phenomenon that are discerned in more complex conceptions. On this basis, conceptual development is seen as requiring the expansion of individuals’ awareness to include aspects of the phenomenon that they did not previously discern.

Awareness of an aspect of a phenomenon is posited as dependent on discerning the potential for variation in that aspect. Marton draws on the concept of colour to illustrate the point (Marton & Booth, 1997). If everything in the world were purple, we would have no notion of colour. It is only through variation in the experience of colour that it may be discerned as an aspect of a phenomenon. Until we experience such variation, the aspect is either taken-for-granted, in the sense that we implicitly presume there will be uniformity in that aspect of the phenomenon, or we do not even notice that that aspect of the phenomenon exists. As a consequence, pedagogical strategies focus on introducing variation in the phenomenon into the learning situation, drawing learners’ attention to different aspects of the phenomenon by varying some aspects whilst keeping others invariant (Marton, 2015).⁵

The claim of a dialectical relationship between discernment of different aspects of a phenomenon and experience of different understandings of the phenomenon is hard to understand in the abstract, so in Åkerlind (2018), I extend the example of colour to better illustrate the relationship:

… one may be aware that the phenomenon of colour consists of the experience of different colours (red, yellow, green, blue, etc.) without being aware that such colours can be grouped into primary and secondary categories, where secondary colours are produced by different combinations of primary colours. And one can be aware that there are primary and secondary categories of colour, without being aware that our perception of different colours is created by different frequencies of light. According to Marton and Booth (1997), these different patterns of awareness and lack of awareness of component parts of a phenomenon lead to different ways of experiencing or understanding the phenomenon as a whole. For example, for someone who understands the phenomenon of colour solely in terms of the experience of red, yellow, green, blue and so on, colour will mean something different than it does for someone who understands colour in terms of both the experience of red, yellow, green, blue and so on, and that this experience of colour is created by different combinations of primary colours, and/or by different frequencies of light. In this way, different understandings of any phenomenon can be traced back to different patterns of awareness of constituent parts of the phenomenon. (p. 950)

Thus, the emphasis in the notion of conceptual expansion is on facilitating the learner to become aware of aspects of the phenomenon that they had not previously discerned, where discernment of a new aspect is dialectically accompanied by a new and more complex way of understanding the phenomenon.

Misunderstanding 8: that phenomenography has stayed the same since its inception in the 1970s

Another common misunderstanding is that phenomenography of the 2020s is the same as phenomenography of the 1980s, when it was first introduced. Too often, I see approaches to phenomenographic research justified solely on the basis of references to Marton’s papers from the 1980s or early 1990s. Whilst Marton’s early papers are still relevant, subsequent developments in phenomenographic research also need to be taken into account.

From the 1990s, additional methodological expectations (Bowden & Walsh, [1994] 2000; Ashworth & Lucas, 2000; Bowden & Green, 2005; Åkerlind, [2005] 2012), and additional theoretical expectations (Marton & Booth, 1997; Marton & Pong, 2007; Pang, 2003), not present in those early papers, were being clarified. I have summarised these developments elsewhere (Åkerlind, 2015, 2018), so will not go into detail here. But it is important to note that, whilst research based solely on Marton’s early papers is still recognisably phenomenographic, it does not have the sophistication and rigour of research that is also based on later works.

Phenomenography: a summary of key distinguishing features

Having spent some time explaining what phenomenography is not, I will now conclude this paper with a summary of what it is. Hopefully, this summary will be more digestible and meaningful after the preceding discussion.

Phenomenography provides a unique approach to exploring human understanding of the world. It can be distinguished from other qualitative research approaches by a combined focus on:

• Qualitative differences – a search for qualitatively distinct differences, not nuanced differences, in ways of understanding the same phenomenon within a sample group;

• Collective understanding – a focus on collective experience, not on individual experience, across the participant group as a whole, with the different understandings constituted in relation to each other, not independently;

• Critical aspects – a focus on critical aspects of experience, not the richness of experience, with the aim of identifying the minimum number of aspects necessary to distinguish one way of understanding from a qualitatively different way;

• Awareness and discernment – an epistemology that positions different understandings of phenomena as based on awareness, or discernment, of different aspects of the phenomena;

• Non-dualist epistemology – a relational epistemology that positions human experience as a person-world relationship;

• Structural relationships – different understandings are seen as inevitably related, through shared awareness of certain aspects of the phenomenon;

• Hierarchical inclusiveness – this shared awareness of certain aspects typically enables the different understandings to be organised in an inclusive hierarchy of expanding awareness of an increasing number of aspects of the phenomenon;

• Part-whole structure (1) – this organisation assumes a part-whole structure to collective understanding, with different ways of understanding the same phenomenon within any group seen as interconnected parts of the whole composite of understandings;

• Part-whole structure (2) – there is also a part-whole structure to human experience at the individual level, with different holistic understandings of a phenomenon seen as constituted from awareness of different aspects or parts of the phenomenon;

• Learning and teaching – there are strong implications for teaching and learning, with the analytic delineation of different aspects and understandings of a phenomenon able to inform the design of learning experiences.

Conclusion

Phenomenography has had a major impact on the field of higher education research and the practice of academic development, especially in the UK, Australasia and Europe. Even though the dominance of its impact has passed, it continues to grow in popularity as an educational research method, forming an ongoing stratum in higher education and academic development research and literature. However, a number of misunderstandings of the approach have developed over time and are continuing to circulate, leading to misinterpretations of the outcomes of phenomenographic research and potential confusion for research students and other researchers interested in phenomenography as a research approach.

Eight common misunderstandings of phenomenographic research were identified in this article:

1. that ‘approaches to learning’ research constitutes phenomenographic research;

2. that the aim of phenomenography is solely to investigate different ways of understanding a phenomenon;

3. that the hierarchical ordering of different understandings in phenomenographic outcomes is based simply on a value judgment;

4. that the interpretive role of the phenomenographic researcher is a subjective one;

5. that the understandings expressed by interviewees in a phenomenographic interview can be taken at face value;

6. that phenomenography investigates individual participants’ ways of understanding;

7. that phenomenography adopts a similar cognitive epistemology to constructivism; and

8. that phenomenography has stayed the same since its inception in the 1970s.

Using the phenomenographic principle of ‘contrast’, each of these misunderstandings has been used in this paper to provide insight into the complex nature of phenomenography, emphasising phenomenography’s focus on:

• Qualitatively distinct differences in human understanding;

• Collective, not individual, understanding;

• Descriptions that highlight only the critical aspects of understanding;

• Awareness and discernment of different aspects of phenomena;

• A non-dualistic epistemology that positions human experience as a person-world relationship;

• Structural relationships between different ways of understanding;

• Hierarchically inclusive organisation of different ways of understanding;

• Part-whole structures to human experience and understanding; and

• Implications for learning and teaching.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 Ference Marton, John Biggs, Paul Ramsden, Noel Entwistle, Michael Prosser, John Bowden and Keith Trigwell. Though it should be noted that Biggs and Entwistle, whilst users of phenomenographic research findings, did not undertake phenomenographic research themselves.

2 This paradigm shift formed part of a larger shift in social science research away from quantitative research and towards qualitative research (Dall’Alba, 1996).

3 Ensuring that all transcripts are able to be ‘allocated’ to one of the categories of description in the outcome space is one way in which phenomenographers check the quality/completeness of their research outcomes, because it is a way of ensuring that there are no transcripts (ways of understanding) unaccounted for in the research outcomes.

4 I believe this to be the reason why the term, conception, which was initially popular for describing the outcomes of phenomenographic research, largely fell out of favour and was replaced in popularity by ‘a way of experiencing’ or ‘a way of understanding’.

5 This has led to a separate strand of pedagogical research, called ‘variation theory research’. Whilst based on the same theoretical assumptions as phenomenography, this research focuses on empirical analysis of practical applications in the classroom (see Åkerlind, 2015, 2018).