As can be seen from the review of existing literature reviews of primary-secondary transitions, only 12 reviews were published between 2008 and 2022; of these none were methodological reviews (Table 1).

Further, none of the 12 reviews had considered the congruence between the philosophical perspectives of the studies’ authors and the research methodology adopted. Similarly, 11 reviews had not focused on the congruence between researcher/s’ conceptualisation of transitions and methodology. The only exception was one review by Jindal-Snape et al. (2020) in which the authors briefly discussed whether the conceptualisation of transitions as an ongoing process would have resulted in more longitudinal studies, with data collection over multiple years and time points. The authors’ highlighted that of the 96 papers (from 2008–2018) included in their review, only 46 had reported on studies that used a longitudinal design, with less than one third of these collecting data at more than three time points and over half at two time points. This lack of focus on congruence is not surprising given that a mapping review undertaken by Jindal-Snape et al. (2021) found that most authors, across the world, did not clearly define the conceptualisation of transitions for their studies, and even those who did, did not make it clear whether that had underpinned their research methodology. None of the 12 reviews focused on the author/s’ theorisation of transitions and its congruence with methodology.

When it came to reviewing the quality and rigour of the methodology of the included studies, only six reviews highlighted them (namely, Topping, 2011; Hughes et al., 2013; Pearson et al., 2017; Evans et al., 2018; Jindal-Snape et al., 2020; Mumford & Birchwood, 2021). Of these, only two undertook and presented a detailed review of included studies (Hughes et al., 2013; Pearson et al., 2017). The other four included one brief generic statement of quality and rigour (Topping, 2011), one critique of data analysis (Mumford & Birchwood, 2021), and two reviews included critique of methodology when presenting the results (Evans et al., 2018; Jindal-Snape et al., 2020). Some reviews, including the four above and Stack et al. (2021), indicated that they had undertaken a quality appraisal, especially when deciding to include the studies or not, but did not present any data related to that.

Hughes et al.’s (2013) review included five papers on primary-secondary transitions of children with ‘special education needs’ published between 2000 and 2011. They developed a quality criteria system to rank the included studies, with total possible scores ranging from 0 to 23, with higher scores indicating higher quality. They further provided detailed critique of the methodology of each study. Pearson et al. (2017) included 10 studies in their review, conducted between 1999 and 2015, with a focus on sedentary behaviours across primary-secondary transitions. They assessed the methodological quality of the reviewed papers using a 9 to 10 item quality scale, which measured four dimensions, study population and participation rate, study attrition, data collection and data analysis. They also undertook independent cross-reviews for each paper and presented the quality scores.

None of the reviews in the last 14 years were methodological reviews (see Table 1). Further, given the paucity of reviews that focussed on the congruence between methodology and philosophical perspective, conceptualisation and theorisation, along with limited robust evaluation of the quality and rigour of the studies, the aim of this study was to undertake a methodological review and mapping of international primary-secondary school transitions literature to ascertain the following:

In this context, by methodology we mean research design, data collection methods and data analysis.

Based on our previous literature review and mapping review (Jindal-Snape et al., 2020; Jindal-Snape et al., 2021), the Evidence for Policy and Practice Information and Co-ordinating Centre’s (EPPI-Centre, 2010) method for undertaking systematic literature reviews was used for the initial search and screening of articles. Subsequently, the Mixed Methods Appraisal Tool (MMAT) (Hong et al., 2018) was used to review the included studies. The key steps of the search and subsequent decisions are outlined in Figure 1.

Of the 151 papers, only 15 papers made explicit reference to the researcher/s’ philosophical perspective and for 13 papers the congruence with methodology was rated as excellent (congruence with every aspect of the methodology, namely research design, data collection methods, data analysis). As can be seen from Table 3, of the 13 papers that were rated to have excellent congruence, 11 employed qualitative approaches, such as, Grounded Theory approach, Qualitative case study, Critical Realist perspective, Participatory action research, Interpretivist paradigm. However, two papers (Bagnall et al., 2021a; Kenyon et al., 2020) stated that they used a quantitative paradigm and quantitative measures respectively (Table 3).

Tobell and O’Donnell (2013) also explicitly stated their philosophical perspective (ethnographic approach), however the congruence with methodological design was rated as good (i.e., congruence with at least two aspects of the methodology). Although Foley et al. (2016) made their philosophical perspective (ethnographic approach) explicit, the congruence with methodological design was rated as unsatisfactory. The reason for this rating was that they reported that ethnographic research formed the basis of this study while employing a mixed methods approach.

In 129 papers (85%) the authors determined the researcher/s’ conceptualisation of transitions. Some of the papers referred to one conceptualisation whereas others had multiple conceptualisations (e.g., Bagnall et al., 2021a referred to Discontinuity due to the multiple changes negotiated, Rites of Passage literature and a Multidimensional process) (Table 4).

In addition to looking at researcher/s’ conceptualisation of transitions the authors were also interested in the congruence between the researcher/s’ conceptualisation of transitions and the methodology.

124 papers, out of 129, (96%) were assessed as being excellent in terms of the congruence between the researcher/s’ transitions conceptualisation and the methodology. This included: research design, data collection methods, data analysis, sampling strategy, sample size, participant group and participants’ views. Of the qualitative studies, 23 were assessed as having excellent congruence (12 had been cross-checked); for quantitative RCT, 8 were assessed as having excellent congruence (6 had been cross-checked); for quantitative non-randomised, 51 were assessed as having excellent congruence (22 had been cross-checked); for quantitative descriptive, 23 were assessed as having excellent congruence (3 had been cross-checked); and for mixed methods, 19 were assessed as having excellent congruence (9 had been cross-checked).

Examples of papers which were assessed as having excellent congruence from each of the study design categories are detailed in Table 5.

Two of the papers were rated as good (Munthe & Thuen, 2009, Neal & Frederickson, 2016); and three were rated as satisfactory (Bunn & Boesley, 2019; Raccanello et al., 2021; Uka & Uka, 2020) in terms of congruence between researcher/s’ conceptualisation of transitions and the methodology.

It was apparent that irrespective of the study design or conceptualisation of transitions there was excellent congruence with the methodological approach.

Of the 151 papers, 86 papers made no reference to researcher/s’ theorisation of transitions, and for 3 papers the researcher/s’ theorisation of transition was unclear.

However, 62 papers did refer to the researcher/s’ theorisation of transitions; of which 14 papers referred to two theories (e.g., Hannah & Topping, 2013), and 3 papers three theories (e.g., Symonds & Hargeaves, 2016). For papers which referred to two theories, reference to Stage Environment Fit Theory (Eccles et al., 1989) and Self-Determination Theory (Deci & Ryan, 1985), was most common, which is in line with theoretical understanding within the field, pertaining to the links between the two theoretical frameworks. Researcher/s’ theorisation of transitions are presented in Table 6.

As shown in Table 6, it was more common for authors to draw on seminal theoretical frameworks to explain the context of school transitions, developed some decades ago and in different educational and cultural contexts. The most popular theorisation being Eccles and Midgley (1989) Stage Environment Fit Theory (n = 17), followed by Bronfenbrenner’s (1979, 1992) Ecological Systems Theory (n = 10), which is in line with Jindal-Snape et al.’s review (2021) findings. In comparison, only two studies referred to more contemporary transitions theorisations such as Jindal-Snape’s (2016) Multiple and Multi-Dimensional Transitions Theory (n = 2).

Furthermore, in comparison to drawing on theoretical frameworks developed within transitions literature, such as Jindal-Snape’s (2016) Multiple and Multi-Dimensional Transitions Theory, many researchers applied theories developed in other domains in different contexts, such as Vygotsky’s (1978) Socio-cultural Perspective on Learning and Pekrun’s (2006) Control Value Theory of Achievement and Emotions. These theories map onto outcomes, or factors associated with school transition, such as attainment (Wheelan et al., 2021), or school belonging (Cueto et al., 2010), as opposed to the transition process.

Through selected theoretical frameworks, there were also theoretical differences in the presentation of school transitions as an ‘event’ (e.g., Lazarus and Folkman’s (1984) Cognitive-Transactional Stress Theory and Coleman’s (1989) Focal Theory of Change), as opposed to a ‘longitudinal process’ of assimilation (e.g., Elder’s (1998) Life Course Theory and Jindal-Snape’s (2016) Multiple and Multi-Dimensional Transitions Theory). This can shape research design decisions, such as the selection and operationalisation of self-report measures, as discussed in Bagnall and Jindal-Snape’s (2023) systematic literature review. This is reflected in the papers which were rated as having excellent congruence with the research methodology and is more problematic for papers which drew on one theoretical framework (e.g., Mackenzie et al., 2012), as opposed to several (Bagnell et al., 2020) demonstrating poor congruence, which is discussed in detail below.

Furthermore, several of the researcher/s’ theorisation of transitions followed a negative discourse, such as Lazarus and Folkman’s (1984) Cognitive-Transactional Stress Theory and Simmons and Blyth’s (1987) Cumulative Stress Model. Drawing on Bagnall and Jindal-Snape’s (2023) systematic literature review findings, and acknowledging that school transitions can be simultaneously exciting and worrying for children (Jindal-Snape, 2016, 2018), this negative discourse is inaccurate; can be leading; and can have negative empirical, practical and political implications. In addition to this, many theoretical frameworks also positioned school transitions as an individual process, which was represented in more seminal theories such as Deci and Ryan’s (1985) Self-Determination Theory. These theoretical frameworks similarly do not holistically capture the nuance of this period, and instead place significant emphasis on the child’s individual characteristics.

In addition to looking at researcher/s’ theorisation of transitions, the congruence between researcher/s’ theorisation of transition and the methodology was also examined. Of the 62 papers which provided a theorisation of transitions, 54 papers were assessed as excellent in terms of their congruence between the researcher/s’ theorisation of transitions and methodology. This included research design, data collection methods and data analysis. Of the qualitative studies, 12 were assessed as excellent (6 had been cross-checked); Quantitative RCT, 4 were assessed as excellent (1 had been cross-checked); Quantitative non-randomised, 22 were assessed as excellent (11 had been cross-checked); Quantitative descriptive, 9 were assessed as excellent (3 had been cross-checked); and Mixed methods, 7 were assessed as excellent (4 had been cross-checked).

Examples of papers which were assessed as excellent from each of the study design categories are detailed in Table 7.

Of the 62 papers which provided a theorisation of transitions, 6 papers were assessed as good in terms of their congruence between the researcher/s’ theorisation of transitions and two aspects of the methodology. For 4 of these papers the aspect of the methodology which was not congruent was the research design, and for the other 2 papers, the aspect of the methodology which was not congruent was within the data collection methods. Of the Qualitative studies, 3 were assessed as good (2 had been cross-checked); Quantitative RCT, 1 was assessed as good (1 had been cross-checked); and Quantitative non-randomised, 2 were assessed as good (1 had been cross-checked).

Of the 62 papers which provided a theorisation of transitions, 2 papers were assessed as satisfactory in terms of their congruence between the researcher/s’ theorisation of transitions and one aspect of the methodology. For one paper, this was the analysis (McCoy et al., 2020) and for the other the research design (Mudaly & Sukhdeo, 2015). Of the Qualitative studies, 1 was assessed as satisfactory (1 had been cross-checked); and Quantitative descriptive, 1 was assessed as satisfactory (1 had been cross-checked).

As noted earlier in the section on quality appraisal, to progress to RQ4, reviewers had to score either RQ 2 and/or 3 as ‘Yes’. Of the 151 included review papers, 6 were not rated in RQ4 as both RQ2 and RQ3 were rated as ‘No’ (Bailey & Baines, 2012; Dæhlen, 2017; Keay et al., 2015; Lester et al., 2019; Pallisera et al., 2016; Schwerdt & West, 2011).

The numbers of included papers in each methodological category are detailed in Table 8.

For each of the categories we evaluated whether each of the 5 criteria were ‘Yes’/’No’/’CT’. The criteria were specific to each category and are detailed in Tables 9, 10, 11, 12, 13.

Using the MMAT critical appraisal tool, primary-secondary transitions studies which adopted a qualitative methodological approach and design were generally found to be of a high standard, with 27 papers (79%) rated as ‘Yes’ for all criteria. In contrast, one study (Mudaly & Sukdeo, 2015) was rated as ‘No’ for four of the five criteria. This study did not discuss the methods adopted and the authors present their analysis and interpretations within the discussion section. In addition, the findings were deemed to be unclear. Two papers (Brewin & Stratham, 2011; Knesting et al, 2008) were rated as ‘CT’ for three of the five criteria (see Table 9). Brewin and Stratham (2011) adopted a longitudinal research design collecting data at two time points from one cohort who had already experienced the move to secondary school. In contrast, Knesting et al. (2008) adopted a longitudinal research design collecting data at three time points in the first year of secondary school from pupils and at two time points from parents and teachers. Thus, a minority of papers have detracted from the overall high standard of qualitative studies. The majority of studies adopted a longitudinal design. However, given the small numbers it would be inappropriate to conclude that adopting a longitudinal design impacts on the overall quality of studies in this category.

The critical appraisal of primary-secondary transitions studies which adopted a quantitative randomised controlled trials methodological approach and design were found to be of a good standing. Three out of eight papers (38%) were rated as ‘Yes’ for all criteria (Pyne & Borman, 2020; Waters et al., 2014a; Waters et al., 2014b). One paper was rated as ‘No’ for one criterion (van Rens et al., 2020). The van Rens et al. (2020) paper indicated a high level of attrition with 371 children (54%) participating in the post transition data collection. One paper (Farmer et al., 2011) was rated as ‘CT’ on all five criteria. This was a longitudinal study with schools randomly assigned to an intervention (teachers received professional development) group and a control group. Thus, a few papers detracted from the overall standard. All studies which were rated as ‘No’ or ‘CT’ on at least one criterion were of a longitudinal design. Again, given the small number of studies, it would be inappropriate to conclude that adopting a longitudinal design impacts on the overall quality of studies in this category.

All 57 papers in this category used measurements appropriate to both the outcome and intervention. Fifty-six of the 57 papers administered the intervention as intended. In one paper it was not possible to determine this (Benner & Wang, 2014). In most papers (n = 44) the participants were representative of the target population. Only 21 papers were rated as ‘Yes’ for having complete outcome data. In contrast, 24 papers were rated as not having a complete data set. In deciding what was rated as a complete outcome data set for longitudinal studies, we used the figure of 10% for up to one year and 25% for more than one year. Most papers (n = 39) were rated as ‘Yes’ for the criterion of confounders being accounted for in the design and analysis. Eight papers were rated as ‘No’ and 10 as ‘CT’ for that criterion. Of the 57 papers, only 14 (25%) were rated as ‘Yes’ for all criteria. It appears that this was due to issues with attrition in longitudinal studies. This is not a surprising finding in transition studies due to the nature of the phenomenon (Jindal-Snape and Cantali, 2019).

Of the 25 quantitative descriptive studies, there was a range of 19–21 papers rated as ‘Yes’ in the five criteria. For criterion 4.2, six papers were rated as ‘No’ as the sample was not considered to be representative of the target population. Three papers were rated as ‘No’ for three criteria, although not for the same criteria (Uka & Uka, 2020; Virtanen et al., 2019; Virtanen et al., 2020). Similarly, McCoy et al. (2020) was rated as ‘CT’ for three criteria. Two papers were rated as ‘No’ in two criteria (Benner et al., 2017; Raccanello et al., 2021). Strnadova and Cumming (2014) was rated as ‘No’ in one criterion and ‘CT’ in another criterion. Twelve papers out of 25 papers (48%) were rated as ‘Yes’ for all five criteria. Six of the 11 papers with at least one criterion rated ‘No’ adopted a longitudinal design. Two of the 5 papers with at least one criterion rated ‘CT’ adopted a longitudinal design.

Of the 21 mixed methods studies, there was a range of 12–19 papers rated as ‘Yes’ in the five criteria. For criterion 5.1 eight papers were rated as ‘No’ as they did not provide an adequate rationale for using a mixed methods design to address the research question. One paper (Finney et al., 2020) was rated as ‘No’ for all five criteria. In addition, one paper (Scanlon et al., 2016) was rated as ‘No’ for four criteria. One paper was rated ‘No’ for two criteria and ‘CT’ for one criterion (MacKenzie et al., 2012). Eight papers (38%) were rated as ‘Yes’ for all five criteria. Five of the 10 papers (50%) with at least one criterion rated ‘No’ adopted a longitudinal design. Three of the 5 papers with at least one criterion rated ‘CT’ adopted a longitudinal design.

To further investigate the relationship between researcher/s’ conceptualisation and theorisation of transitions and the assessed quality of papers, the authors analysed the number of papers where RQ2 was rated as ‘Yes’ and RQ3 was rated as ‘No’ or ‘CT’ (n = 73); where RQ3 was rated as ‘Yes’ and RQ2 was rated as ‘No’ or ‘CT’ (n = 6); and where RQ2 and RQ3 were rated as ‘Yes’ (n = 56). Thereafter, the authors calculated the number of papers where all 5 MMAT criteria were ‘Yes’. Based on findings, where RQ2 was rated as ‘Yes’ and RQ3 was rated as ‘No’ or ‘CT’, the proportion was 40% (29 out of 73). In addition, where RQ3 was rated as ‘Yes’ and RQ2 was rated as ‘No’ or ‘CT’, the proportion was 33% (2 out of 6). Finally, where RQ2 and RQ3 were rated as ‘Yes’, the proportion was 52% (29 out of 56).