Why assessment in physical education is still problematic: A critical interpretive synthesis of physical education assessment literature

Formulating open research questions

After surveying reviews of literature and conceptual texts (i.e. the literature presented in the Introduction and Background sections above), we developed two “compass questions” (Bullock et al., 2021): (1) How does assessment become problematic in PE? and (2) What solutions do PE scholars offer for addressing the problems they identify? These questions allowed us to narrow our focus in our search for literature and guided our extraction of data once we had found the literature for review.

Searching for literature with a broad strategy

We adopted two search strategies to identify relevant literature. First, we identified texts using the databases Google Scholar and Education Resources Information Center (ERIC). The search combinations employed were “physical education” AND “assessment” and “PE” AND “assessment,” which we used “anywhere in the article.” Second, we relied on snowball retrieval, where we searched reference lists from the texts found with the first strategy to identify further texts. In line with CIS methodology (Dixon-Woods et al., 2006), our intention with this broad search was to find potentially relevant literature from which we could make a more precise selection based on relevance.

Selecting literature based on relevance

To be included in the review, texts needed to: (1) Have been peer-reviewed and published in English; and (2) Have been published between 1999 and 2024. The start year reflects our recognition that both assessment and PE have been affected by societal changes (e.g. increased use of digital technologies) since the turn of the millennium. Our intention with the time span was to be comprehensive but current; (3) Refer to problems, difficulties, or disadvantages with enacting formative and/or summative assessment in school PE. These problems could be described from teachers’, preservice teachers’, students’, researchers’ or other stakeholders’ points of view. Problems did not need to constitute the analytic focus of the study, but assessment did. In other words, if the text referred to assessment but focused on physical literacy (e.g. Jean de Dieu and Zhou, 2021) or fundamental movement skills (Lander et al., 2015) or teachers’ content knowledge (e.g. Ward et al., 2015), for example, the article was not included; (4) Describe a study that involved producing empirical material. Employing this last criterion was based on our intention to contextualize the problems identified.

Evaluating the quality of the literature

We used Dixon-Woods et al.’s (2006: 4) “appraisal prompts for informing judgements about the quality of papers” (see the Appendix) to decide whether a text should be included in the review. In short, if we had concerns regarding the clarity or appropriateness of the study's aims, objectives, research design, methodological approach, and/or interpretation, we sent the text to the other two authors for cross-checking. If, after discussion, we agreed that parts of the text were unclear or aspects of the investigation had been conducted inappropriately, we deemed it to be of insufficient quality and excluded the text from our sample (see Figure 1 for a summary of the search and selection process). Details of the 48 reviewed texts are included in the supplemental material available online. They have been assigned a number (1–48) and are referenced in this article using that number.

OPEN IN VIEWER

Figure 1.

Flowchart of literature search and selection process.

Extracting data from the literature

Unlike in traditional systematic reviews (Bores-García et al., 2020; Otero-Saborido et al., 2021), we were interested in extracting a relatively narrow range of data from the literature, namely the type of assessment in focus, the country, the school level, the reasons for how and why formative and summative assessments become problematic in PE, and the solutions that scholars recommended. To extract this information, we developed a template that we used to record details from each article (see Table 1). Relevant parts of the text were either paraphrased or copy-pasted into the template so that each article had its own Word document. Following Depraetere et al. (2021), we initially systematized the process by completing data extraction individually with the same five articles and then compared our results. Reaching an agreement on the type of assessment in focus, country, and school level was unproblematic. Questions did arise around the identification of problems and solutions so in line with other CIS research (Dixon-Woods et al., 2006), this aspect of data extraction was completed in an inclusive manner. This meant that if there was any question of whether text from the article should be interpreted as describing a problem or solution, it was included and compared against other completed templates later in the process. When templates for all studies were completed, the first author printed all templates, grouped them by assessment type (formative, summative or “formative and summative”), and compared the content of the templates for patterns and similarities. Using pen and paper, the first author categorized problems and solutions under tentative headings such as “teachers’ personal/internalized judgments” and “students’ inexperience.” The headings and placement of example text under the headings were discussed and finalized at a data analysis meeting.

Formulating a synthesizing argument

The final activity entailed creating a synthesizing argument (Farias and Laliberte-Rudman, 2016). Formulating this argument involved critically examining the summary of the extracted data and considering the assumptions that make the problems identified persistent and the solutions ineffective or difficult to enact in practice. The first author started the formulation process by constructing a draft thesis in response to the questions above. Once this draft had been completed, the second and third authors refined the text, adding, deleting and commenting where they saw necessary based on their familiarity with the literature (see Bullock et al., 2021). In this way, the thesis took form as a synthesis of the problems and potential solutions associated with assessment in PE.

Problems with summative assessment

Problems identified with summative assessment can be divided into three broad “areas,” each of which contains related subproblems. These problematic areas overlap but cluster around: (1) the intrusion of teachers’ personal/internalized criteria into the assessment process, (2) students’ negative experiences of assessment, and (3) a lack of curricular clarity regarding how assessment should be conducted.

Teacher use of personal criteria is framed as a problem in that PE teachers collect data for the purposes of providing accounts of student learning without basing the data on criteria that reflect learning (11, 37). Many investigations suggest that teachers use behavioral, dispositional and/or physical factors to assess their students. These factors include whether students: have the correct clothing and participate in lessons (11, 24, 47); are fit and/or appear in good shape (33, 34); are willing to try new activities (39, 46); and demonstrate competitiveness or aggression during games (31). Some studies further suggest that teachers are unaware of the impact their internalized criteria have on their assessment practices (2, 39). Beyond the contradiction of using criteria that do not reflect learning to create accounts of learning—a contradiction that calls into question whether PE teachers are engaging in summative assessment at all when they rely on such criteria—the reliance on personal criteria has been problematized in three ways. Scholars have observed that personal criteria fall outside curricular prescriptions (7), thus creating instructional misalignment between learning goals and assessment practices (11, 26). Further, scholars have pointed out that assessments based on teachers’ internalized criteria lack transparency. Several investigations highlight the gap between the criteria teachers use and the criteria on which students believe they are being assessed (27, 31, 48). Lastly, some scholars have noted that when teachers use their own subjective criteria, it becomes difficult to compare what students have learned across schools or even across classes at the same school (17).

The second problematic area concerns students’ negative experiences of assessment. Negative experiences comprise experiences of assessment as unfair and experiences of assessment as anxiety-inducing. In Røset et al.’s (2023) (34) investigation, boys and girls both suggested that the opposite sex were advantaged in PE assessment (see also 23 for statistics regarding gender and PE grades), while in Modell and Gerdin's (2022) (26) research, some students believed that it was not specifically gender but certain physical characteristics such as strength and speed that were advantaged in PE. In other investigations, students claimed that assessment was unfair because they were assessed on skills and knowledge they had not had the opportunity to learn during lessons (26, 27, 32). One large-scale statistical analysis of grades supports this contention, indicating that prior sporting experience is likely to result in higher grades (38). Noteworthy with respect to equity is that research suggests that other factors such as ethnicity and parents’ educational level (38) and relative age (33) also influence grades, yet students have not associated these factors with assessment inequity.

Another less commonly identified subproblem concerning students’ negative experiences of assessment involved stress (13). Students in Røset et al.'s (2023) (34) investigation claimed that the highly visible ways in which assessment tasks were conducted in Norwegian PE lessons led to anxiety and worry, and feelings of being surveilled. Kim and Lee (2022) (13) observed that, despite experiencing highly stressful assessment practices as school students, PE teachers were inclined to engage in those same practices when teaching.

Significant with respect to students’ negative experiences is that few studies have indicated that students cite lack of transparency as a problem (26 constitutes an exception). This is noteworthy in light of the abundance of research problematizing teachers’ personalized assessment practices (11, 24, 33, 34, 47). One investigation even indicated that students feel relatively confident in recognizing the criteria on which they are being assessed (32). This confidence may be misplaced, although it is also possible that students recognize and react to their teachers’ use of personal, “unofficial” criteria for assessment.

The third problematic area concerns direction provided by state curricula in matters of assessment. On one hand, Spanish researchers have claimed that curricula are becoming closed and prescriptive, and the State is emphasizing standardization and measurement (29). On the other hand, (mainly Scandinavian) researchers have contended that educational policy is too open and has provided teachers with insufficient guidance and direction (15, 32). Svennberg (2017) (37) claimed that national curricula do not provide clear distinctions between grade descriptions and Svennberg et al. (2014: 211) (39) asserted that “as long as the grading criteria do not address classroom realities and restricting factors, the grades will continue to be inconsistent with the [curricular] criteria.” In an investigation that provides some insights into why educational policy might fall short of teachers’ expectations, Scanlon et al. (2019) (35) observed how different stakeholders influenced the construction of the Irish curriculum. The researchers concluded that teachers had markedly different expectations to policy makers and that tensions resulted in compromises for all parties.

Potential solutions to problems with summative assessment

Although researchers have critiqued many aspects of summative assessment, the logic underpinning solutions is remarkably consistent across the corpus of literature. The principal solution is to provide teachers with training either through (better) initial teacher education (7, 25) or, more commonly, through continuing professional development (CPD—24, 31). More precisely, researchers have suggested that teachers need to learn to be more reflective so they can “see,” and thus prevent, their perceptions from “contaminating” (11: 161) their assessment of student learning (39). ³ Teachers need to develop a sound understanding of curricular intentions so they are familiar with what students are supposed to learn (24, 32). Once they understand curricular intentions, they need to ensure that their teaching and assessment practices are aligned with those intentions (7, 46). Redelius and Hay (2009) (31) suggest that to ensure this alignment, teachers must learn to write assessment statements, to collect evidence, and to understand and use this evidence. In the classroom, teachers must be able to communicate learning intentions and assessment criteria clearly to students (11, 27, 37). Zhu (2015: 419) (48) suggests, for example, that PE teachers “should explicitly teach the achievement standards and grade expectations [to students].” While scholars have placed considerable confidence in CPD to improve assessment practices (11, 31), the responsibility of education agencies to provide teachers with clear guidelines that explain how assessment should be carried out has also been cited (11, 32).

Other potential solutions offer alternative ways of understanding the antecedents of problems associated with summative assessment. Marmeleira et al. (2020) (23), for example, propose that boys receive higher grades than girls in PE because girls have lower levels of sport participation outside of school and are thus less familiar with the content dealt with in PE. Their solutions revolve around making PE more accommodating for girls by introducing, for example, single-sex competition and noncompetitive environments. In their discussion of tensions around curricular development for senior PE, Scanlon et al. (2019) (35) note that there is a need for stakeholders to work democratically to determine what should be assessed. And in response to the claim that teachers’ personal criteria for “good performance” contaminate the assessment process, Svennberg et al. (2018) (40) offer an alternative to CPD when they contend that it is impossible to separate behavioral aspects such as attendance and being on-task from knowledge and skills. More specifically, they suggest that it is not possible to comment on students’ development of knowledge if they are not in class or are not attempting to demonstrate the knowledge being learned during lessons. In this respect, they recommend accepting a relationship between behavioral aspects and learning.

Problems with formative assessment

Problems identified with formative assessment can also be divided into three broad “areas,” each of which contains a collection of subproblems. These areas revolve around: (1) teachers’ lack of competence with formative assessment practices; (2) challenges that teachers experience with formative assessment; and (3) problems that students experience with formative assessment.

Scholars have identified many shortcomings with teachers’ formative assessment practices, often referring to teachers’ lack of assessment literacy (9, 17). Slingerland et al. (2024) (36) claimed that teachers lacked proficiency in developing learning targets or success criteria, rarely provided students with information on how to improve their performance and were sometimes unsure of what assessment strategy they were using. Leirhaug and Annerstedt (2016) (15) proposed that the teachers in their investigation constantly provided students with feedback, but that the comments did not relate to specific learning goals, while Chng and Lund (2019) (8) suggested that the teacher in their study failed to place much emphasis on peer feedback. ⁴ Tolgfors (2018) (41) pointed to considerable variation between teachers, suggesting that assessment for learning took different forms in PE, and that often the learning motive for assessment was overpowered by the accountability motive (essentially turning assessment intended for learning into assessment for grade generation—see also 44). Raising the question of teachers’ ability to align different forms of assessment, Leirhaug (2016) (14) observed an apparent paradox, asserting that better performance in formative assessment tasks was not associated with better performance in summative assessment tasks. One study found that a quarter of the 210 PE teachers surveyed claimed not to use formative assessment strategies at all (25).

Given reports of teachers’ lack of assessment knowledge, it is perhaps unsurprising that teachers have expressed apprehension about conducting formative assessment. Some teachers’ concerns are reflected in their reluctance to deviate from traditional PE practices (9). Teachers are worried that more time dedicated to formative assessment will result in less time for physical activity (36). Indeed, teachers in many studies expressed uneasiness at the amount of time taken up by formative assessment (4, 9, 10, 19, 22, 45). In addition to concerns about time, teachers in Slingerland et al.’s (2024) (36) investigation were worried that the increased student autonomy involved in the formative assessment tasks offered them less control and could raise safety issues when done in the context of gymnastics (see also 21). Other studies indicated that teachers lacked confidence either in their own assessment capacities (9) and/or in their students’ capacities to engage in assessment practices (36). Tolgfors et al. (2022) (43) noted that even if teachers have developed a sound understanding of formative assessment principles in initial teacher education and intend to practice formative assessment once they are in schools, they may find it difficult to implement when faced with oppositional expectations from colleagues and students.

A third problematic area relates to students and their lack of experience with formative assessment. Some students in Leirhaug and Annerstedt's (2016) (15) research were unsure of whether they had engaged in formative assessment and claimed to have minimal experience with self- or peer-assessment in PE (see also 20). Students in Butler and Hodge's (2001) (6) investigation reported that they received general feedback from their peers but not specific feedback when working with assessment for learning. In a similar vein, Chng and Lund (2019) (8) observed pupils rushing to complete assessments so they could continue playing games or leave PE for the next class, while Slingerland et al. (2024) (36) noted that the students they observed were simply not particularly good at giving or receiving feedback. Regarding feedback, Backman et al. (2024: 11) (3) identified a slightly different problem, pointing out that students’ physical capital affected “who is allowed to give feedback to whom.” Their results suggest that students with higher status have more to say in peer assessment situations. In this respect, Backman et al.’s (2024) (3) research suggests that, even if students are familiar with technical aspects of formative assessment, interpersonal issues still need to be managed (see also 15 for reference to student marginalization).

Potential solutions to problems with formative assessment

Unlike the solutions offered to improve summative assessment, scholars’ recommendations for improving formative assessment are diverse. Some scholars have adopted a philosophical position, suggesting that tensions will always make formative assessment complex. Tolgfors (2019) (42), for example, draws attention to the influence of sociocultural factors, claiming that assessment practices need to be adapted to the context in which they are taking place. In a separate paper, Tolgfors (2018) (41) suggested that teachers need to employ different versions of formative assessment and that there is no standardized way of doing formative assessment that will be effective in all situations (see also 28, 43 and 44).

Other scholars have assumed more practically oriented positions. As well as broad calls for greater student involvement (4, 16, 22, 42), researchers have made an array of concrete suggestions that are supposed to help students work with formative assessment (1, 8). Lorente-Catalán and Kirk (2016) (18), for instance, suggest that teachers should introduce peer-assessment before self-assessment, reasoning that self-assessment is more difficult and engaging in peer-assessment initially will give students confidence. The same researchers suggest that teachers should give feedback prior to awarding grades in order to place greater emphasis on learning and reduce the emphasis on grades. They further propose that students should be provided with assessment criteria when they are doing self-assessment (18). Fletcher et al. (2024) (9) and Butler and Hodge (2001) (6) provide similar practical recommendations for teachers that include creating safe and trusting environments, using technology to make assessment less time consuming and easier to manage, and maintaining tight control of situations where students are assessing one another. Barrientos Hernán et al. (2023) (4) reported on the practical solutions developed by the teachers in their investigation. These included recording information on a few students in every lesson and using time spent moving between classrooms to hold personal assessment interviews with students and thus protect active learning time in lessons.

As with summative assessment, scholars have underscored the need for teachers to learn more about formative assessment through CPD (4, 18). Slingerland et al. (2024) (36) claimed that, while teachers may resolve some problems in the course of using formative assessment, professional development has the potential to challenge teachers’ fundamental understandings of assessment and provide important opportunities to consider how formative assessment can work in their particular contexts. They outline many characteristics of effective CPD, suggesting that it should involve a “hands-on” approach where teachers can work collaboratively and/or in communities, and focus on content that is aligned with students’ needs. Supporting Slingerland et al.'s (2024) (36) claim, Ní Chróinín and Cosgrave (2013) (28) reported that attending postgraduate studies in PE radically altered primary school teachers’ understandings and practices of formative assessment. Ní Chróinín and Cosgrave (2013) emphasized the value of scaffolding effective assessment strategies, concluding that initial teacher education needs to place more emphasis on assessment and sample assessment tools, and that teachers then need to be supported through CPD. ⁵