Effects of rubrics on learning and performance

Eleven of the 13 studies that investigated effects of rubrics on learning and performance showed that using rubrics was associated with improved learning or higher performance. This section, therefore, does not focus on the point that rubrics can improve performance, which hardly needs further support (^{Panadero et al., 2023}), but rather on sharing descriptions of the training, activities, or instruction that accompanied the rubric use. In other words, if rubrics are tools, what were these tools used to do?

In all 11 studies showing positive outcomes for learning and performance, students were introduced to rubrics and given opportunities to use them. The process familiarized students with the criteria and performance level descriptions and, therefore, helped them form a concept of what good work looks like. In all 11 studies, the rubrics were used formatively. Ten of those studies explicitly described formative uses: as a framework for instruction (^{Bradford et al., 2016}); as the basis of individual and group activities (^{Auxtero & Callaman, 2021}); and/or as the basis of feedback from teacher, peer, or self (^{Chen & Andrade, 2018}; ^{Chen et al., 2017}; ^{Hsia et al., 2016}; ^{Kim, 2019}; ^{King et al., 2016}; ^{Liu et al., 2016}; ^{Safadi, 2017}; ^{Safadi & Saadi, 2021}). One study (^{Mahmood & Jacobo, 2019}) focused on the use of sliding scale rubrics for grading, but the students used the rubrics formatively as they were preparing mathematics portfolios to submit for grading; in other words, they self-assessed as they prepared their portfolios. In this study, “sliding scale” does not mean that the rubric performance level descriptions changed, but that students only saw the descriptions for the levels above and below their current performance. In several studies, it was explicitly concluded, either by author inference or by talking with students, that rubrics helped students understand the teacher’s expectations and/or the elements or components of a particular task. In either case, their attention was focused on the aspects of the task that give evidence of learning (^{Auxtero & Callaman, 2021}; ^{Bradford et al., 2016}; ^{Chen et al., 2017}; ^{Chen & Andrade, 2018}; ^{Kim, 2019}).

As the previous paragraph shows, most of the studies that employed rubrics successfully to increase student learning and performance used the rubrics formatively, to support learning while it was happening. Two of the studies (^{Chen & Andrade, 2018}; ^{Chen et al., 2017}), using a rigorous design, investigated the effects of implementing criteria-referenced tools (whether rubrics, checklists, or something else) accompanied by teacher, peer, or self-assessment that yielded formative feedback to students and provided the opportunity for revision of work. This combination led to a positive effect on learning and performance in the arts compared with a control group. In this intervention, the salient characteristics of the treatment were providing criteria and ensuring that students used them - often for selfor peer-assessment but also, and certainly, for improving their work. This combination of criteria and certain student use turns out to be a thread through all of the studies that reported improved learning and performance.

Using rubrics for peer-assessment (peer feedback) figures in two more of the successful studies, both in the context of web-based learning. ^{Hsia et al. (2016)}, using a quasi-experimental design, studied web-based assessment in a junior high school performing arts course where the students wrote and videotaped short dramas based on Chinese folk stories. Both groups were given the rubric, but only one group participated in a peer review process. The control group also had the rubric but there was no guarantee any of the students used it to view their group’s video performances, as the treatment group had to do. ^{Liu et al. (2016)}, also using a quasi-experimental design, investigated the effects of peer review using a checklist of five criteria presented in sentence form (e.g., “The story has a vivid background, events, actions, and ending,” p. 289) on an iPad storytelling activity called “Saving the Forest”. The control group could view and discuss the stories of others but did not use the criteria. Again, the common thread seems to be clear presentation of criteria linked to some activity that ensured students used the criteria.

Four other studies displayed this combination of presenting criteria clearly and ensuring students used the criteria purposefully. Two of those studies were descriptive (no control group) and used true rubrics. ^{Kim (2019)} found that high school students who used rubric-referenced self-assessment wrote essays that were longer and of higher quality after using the rubric. ^{King et al. (2016)} found that co-creating a rubric for writing mathematical justifications, and then using their rubric for self-assessment that included both feedback to themselves and self-scoring, helped 5^th graders distinguish between simply presenting the steps to solve a math problem and explaining the reasoning behind those steps, which had been an issue for them before the lessons that used the rubric.

Two additional studies from one research group (^{Safadi, 2017}; ^{Safadi & Saadi, 2021}) used a rigorous design to test the effectiveness of an instrument they called a rubric that supported a self-diagnosis activity. After solving force and motion problems (^{Safadi, 2017}) or geometric optics problems (^{Safadi & Saadi, 2021}), students in the treatment group reviewed a complete, worked example for each of the problems, along with instructions to assign themselves a score, according to directions, for each step in the problem. Students in the control group experienced whole-class discussion of the worked examples without the self-diagnosis and scoring activity. Again, what was tested was the clear presentation of criteria and an activity that ensured the students needed to engage with those criteria.

Only 2 studies that investigated the effects of rubrics on learning and performance did not find positive results, and it is not possible to generalize from the results of these studies. However, it is worth noting that, in one of those studies (^{Hubber et al., 2022}), students were given a rubric only during the final week of a three-week interdisciplinary science and mathematics unit, when their culminating group project (the marble run challenge) was already in progress. By then, the students were past the design phase and into the construction and testing phase of their project constructing marble runs. ^{Hubber et al. (2022}, p. 16): “A rubric . . . was introduced in the final week and students were asked to review their work against the rubric and teachers highlighted the need for students to self-assess their work and the teamwork of the group.” No guided activities were described for the students, so presumably the students’ use of the rubric could have been perfunctory, or they could even have ignored the rubric.

In the other study, that was also not able to establish an association between rubric use and improved performance (^{Smit et al., 2017}), students did participate in selfand peer-assessment. The use of rubrics had direct effects on their perceptions of formative feedback and self-assessment, and indirect effects on self-regulation and self-efficacy. The authors suggested that these effects might have mediating effects on outcomes, that is, improvement in self-assessment might ultimately lead to improved learning and performance.

Effects of rubrics on student motivation

Three of the 4 studies that looked at the association between rubric use and student motivation found positive outcomes (^{Hsia et al., 2016}; Nsabayezu, 2022; ^{Smit it al., 2017}), and the fourth found no difference (^{Liu et al., 2016}). The 3 studies stating positive motivational outcomes found rubric use was associated with increased student self-efficacy specific to rubric use. Hsia et al. (2017) showed in a quasi-experimental study that for students who participated in online peer-assessment, self-efficacy for peer-assessment, especially for evaluating peers’ work and for receiving peers’ feedback, was related to performance. ^{Smit et al. (2017)} used a quasi-experimental design and showed, with causal modeling, that rubric use had a direct effect on students’ perceptions of their peerand self-assessment skills and formative feedback, as well as indirect effects on both self-efficacy and self-regulation. In a descriptive case study, Nsabayezu (2022) reported students expressed general satisfaction and motivation for rubric use.

The one study that found no difference in self-efficacy, between students who used rubrics and those who did not, used peer-assessment with checklist-style items about creating stories using iPad. The study measured creative self-efficacy, defined as students’ confidence in creating novel works (^{Liu et al., 2016}, p. 287) and not as self-efficacy for using rubrics. While the peer review group got significantly higher scores on their stories, on almost every dimension of the rubric, there was no significant difference in creative self-efficacy between the peer review group and the control group. It is worth pointing out that creative self-efficacy is a broader construct than self-efficacy for using rubrics, and may require more than one encounter to change.

Effects of rubrics on student attitudes and perceptions

Student attitudes and perceptions affect their engagement with any instructional tool or strategy. Four studies reported on student attitudes and perceptions, but the focus of the attitude questions differed (e.g., attitude toward the content taught, attitude toward rubrics, or both). In a quasi-experimental study, ^{Bradford et al. (2016)} found no difference between the rubric and non-rubric groups on attitudes toward writing, but the primary (grades 1-2) students’ written reasons for their attitude ratings revealed they thought using rubrics allowed them to finish their opinion paragraphs quickly and think they performed well in some writing skill.

On balance, in descriptive (non-experimental) studies students reported a generally positive view of rubrics and some negative attitudes, as well. Teachers in Gallego-Arrufat and Dandis’s (2014) study said using a rubric enhanced secondary students’ engagement and learning, as well as their own teaching; but, that the lack of experience with rubrics, the length and complexity of the rubric, and the amount of time needed to use this new tool created difficulties for both teachers and students. The authors reported some resistance regarding the subject matter as these students were in their third semester, had never done mathematical explanations before, and some complained about having to learn this new content. ^{Kim (2019)} found secondary students came to perceive that rubric-referenced self-assessment was effective and that it positively affected their attitudes toward writing. Also working with secondary students, ^{Mahmood and Jacobo (2019)} found that using sliding-scale rubrics led most students to report positive attitudes and motivation to improve, but some students were hesitant.

Teacher use of rubrics

Findings about teacher attitudes, opinions, and experiences with the use of rubrics do not speak directly to student outcomes, but they do speak directly to implementing rubrics in the classroom. Implementation is a necessary step before any educational tool or strategy can have an effect on students and their learning. Three studies collected interview and diary (^{Gallego-Arrufat & Dandis, 2014}) data from teachers or their coaches (^{Kennedy & Shiel, 2022}), or students (^{Idris et al., 2017}) about the use of rubrics in the classroom.

Teachers use rubrics to clarify expectations (^{Gallego-Arrufat & Dandis, 2014}; ^{Idris et al., 2017}; ^{Kennedy & Shiel, 2022}). When rubrics are clear, students do use those rubrics to understand what is expected of them (^{Gallego-Arrufat & Dandis, 2014}; ^{Idris et al., 2017}; ^{Kennedy & Shiel, 2022}). Teachers also use rubrics in formative assessment, for example: for giving feedback for selfand peer-assessment (^{Idris et al., 2017}; ^{Kennedy & Shiel, 2022}), for increasing the objectivity of summative evaluation (grading; ^{Gallego-Arrufat & Dandis, 2014}), for promoting student thinking skills (^{Idris et al., 2017}), and for planning their lessons (^{Gallego-Arrufat & Dandis, 2014}; ^{Kennedy & Shiel, 2022}). If the rubric is designed to support it, teachers can use rubrics to identify students’ developmental level on a skill like writing (^{Kennedy & Shiel, 2022}).

Two main difficulties or obstacles to using rubrics were reported. One was the amount of time it took to learn how to use rubrics, for those for whom they were new (^{Gallego-Arrufat & Dandis, 2014}). A second obstacle was that the design of the rubric and the language needed to be understandable to students (^{Gallego-Arrufat & Dandis, 2014}).

Ensure that rubrics are of high quality

Rubrics that are of high quality are based on clear, appropriate criteria that are drawn from the learning goal they are meant to assess, not the task (^{Andrade, 2000}; ^{Brookhart, 2013}; ^{McTighe & Frontier, 2022}). For example, in a writing rubric, the criteria ideas, organization, word choice, sentence fluency, conventions, and voice (^{Bradford et al., 2016}) are derived from the characteristics of effective writing that the students are to be learning, not from the characteristics of the task (e.g., the rubric did not include a criterion like “took a position about the best kind of pizza”). Such criteria identify evidence of learning across a range of similar tasks (e.g., an opinion piece about the student’s favorite subject in school, or about the best flavor of ice cream). This, in turn, helps clarify the learning goal for students, give meaning to the task, and focus students on improving (in this case) their writing rather than completing one assignment.

Rubrics that are of high quality feature performance-level descriptions that cover the continuum of work quality from low to high. Even if a teacher does not expect any student work to be located at a particular level, for example if the teacher does not think that any work will be of the lowest quality level on the performance scale, the description should still be there to communicate the range of performance and to give students a sense of where on that range their current performance falls. Teachers can help students review illustrative examples of work at the various levels to help them develop a concept of what the criterion means (^{Andrade, 2000}; ^{Brookhart, 2013}). Performance level descriptions should be understandable to the students and teachers who will use them; otherwise, they will not be useful. In pursuit of this goal, use student-friendly language - employ both words and phrases that students can hear themselves using - and use nouns often instead of pronouns (e.g., “my organization,” not “it”).

Actively engage students with the rubrics

Successful studies in this review used various strategies to actively engage students with rubrics, including: giving students practice using the rubrics individually or in groups, on prepared work samples; basing teacher feedback on the criteria and providing students opportunities for revision; involving students in rubric-referenced selfor peer-assessment; and, using rubrics as a framework for planning instruction. These active strategies are consistent with other recommendations (e.g., ^{Andrade, 2000}; ^{Brookhart, 2013}; ^{McTighe & Frontier, 2022}; ^{Panadero et al., 2016}).

Simply distributing a rubric among students does not guarantee they will read it, much less understand it. ^{Auxtero and Callaman (2021)} planned a group activity in which students in the treatment group were familiarized with the rubric, then an activity was done in pairs for additional practice using the rubric. Other ways to introduce rubrics to students include having them ask questions, having them rewrite the performance level descriptions in their own words, or having them sort and match sample student work (or mocked-up sample work) to the performance level descriptions for each criterion. For learning goals with which students are already somewhat familiar (e.g., mathematical justification, ^{King et al., 2016}), students can co-create the rubrics with the teacher.

Once the rubric is familiar to the students and their conception of the learning goal is beginning to be shaped by it, students can use the rubric for peeror self-assessment, as was done in all of the successful studies in the present review. Students will need instruction on how to match qualities in the work with qualities described in the rubric. One strategy is to use a highlighter to identify the evidence in the work with the same color as the description in the rubric, perhaps working in pairs. For example, if a mathematical problem-solving rubric specifies that the student should write an equation modelling the problem, then the students would highlight the equation in the work and “write an equation modelling the problem” in the rubric with the same color.

For self-assessment, students match the descriptions in the rubric with their own work. Then, they use the description in the next level of the rubric to set a goal for improvement. First, the teacher should make sure the students understand the purpose of the self-assessment, that it is not self-grading but rather reviewing their work in order to make it better. Give students an activity or protocol, for example, using a copy of the rubric with a comments section beside it to make notes. However the self-assessment is structured, students also need an opportunity to use their self-assessment feedback to revise their work (^{Chen & Andrade, 2018}; ^{Chen et al., 2017}). Self-assessment feedback is moot if it is not used, and students will soon see it as futile. In contrast, closing the loop by seeing the self-assessment leading to improvement will contribute to students’ self-regulation of learning (^{Panadero et al., 2016}). Give students feedback on the quality of their self-assessment and make sure to explicitly connect the self-assessment they did with the improvement in their learning.

Peer-assessment is similar to self-assessment in that students match descriptions in the rubric with work, but they do this for peers. In addition to clarifying the purpose for the peer-assessment and using clear rubrics, peer-assessment requires some attention to the social nature of this learning strategy. Match the participants into compatible pairs or small groups that can work together. Make sure that students understand they are to assess the work and not the person and, if writing is involved in the peer feedback, to make it clear and helpful to the peer who will receive it. As with self-assessment, give students feedback on the quality of their peer-assessment, give peers an opportunity to revise their work, and make sure to explicitly connect the improvement they see with the peer-assessment they did.

Teachers can use the criteria in a rubric as a framework for planning their instruction in lessons or mini-lessons (^{Gallego-Arrufat & Dandis, 2014}; ^{Idris et al., 2017}; ^{Kennedy & Shiel, 2022}; ^{Smit et al., 2017}). This author has observed lessons where teachers structured their didactic introduction to a lesson or topic by illustrating the criteria and then used the language of the criteria, as they observed and commented, while students were working on individual or group activities. The likely reason that structuring a lesson around the criteria is effective is that, if the criteria are appropriately linked to the learning goal and not the task, student understanding (or skill in demonstrating) the criteria is equivalent to students achieving the learning goal.

Use rubrics to connect formative assessment and grading

Only one study in the present review focused on rubrics used for grading (^{Mahmood & Jacobo, 2019}). It is worth noting that in that study, students used the rubric formatively, self-assessing their mathematics portfolios against the rubrics with which they would be graded. The fact that rubrics can make this connection between formative assessment and summative assessment (grading) is one of the benefits of using rubrics in the classroom, and has been noted before (^{Brookhart, 2013}). There are several ways to connect the criteria used in rubrics-referenced formative assessment to the criteria ultimately used for grading. Sometimes, lessons tackle one criterion at a time. For example, a writing teacher might first focus on the ideas in students’ work, then on word choice, and so on, culminating in the full use of the criteria at the end of a unit. Or, a music teacher might focus on students singing the correct pitch first, then on tempo, then dynamics, culminating in the full use of the criteria at the final performance.

Another way to connect the criteria used in rubrics-referenced formative assessment to the criteria ultimately used for grading is to use “lesson-sized” versions of the criteria for initial instruction. For example, in a primary school lesson, students might focus on the use of periods at the end of sentences, then on the use of question marks at the end of questions and, finally, on the use of exclamation points at the end of exclamations, culminating in a rubric which says, in part, something like, “I put the right end punctuation at the ends of my sentences”.