Comprehensive Assessment Research Review: Providing Feedback

Clear criteria empower learners to be self-directed and provide a tool for gathering actionable feedback throughout the learning process. This section describes a range of research-based tactics to elicit evidence of learners’ understanding and provide feedback that promotes learning.

Assessing learning throughout the process is akin to checking one’s location periodically along a journey to ensure one’s timely arrival at a destination -- as opposed to checking only at the end to see how far off course one has drifted (Wiliam, 2010). Quality of learning can be observed through what students say, write, make, or do (Griffin, 2007). To ensure learning stays on course, feedback must (a) be based on evidence of student understanding and integrated into regular classroom practice, (b) yield actionable insights that learners can and will use, and (c) provide learners with opportunities to improve their performance (Shute, 2008).

Rather than simply indicating correct or incorrect, feedback should give students enough information to enable them to correct errors by themselves (Shute, 2008). It should focus on the what, how, and why of a problem and provide suggestions about what the learner can do to improve specific qualities of their work in relation to specific features of the task (Shute, 2008). Feedback is less effective when it focuses on the self or makes comparisons with other pupils; it is more effective when it focuses on the task details, methods to improve answers, and ways to set appropriate goals (Shute, 2008).

Feedback can be either delayed or immediate relative to the particular task and to the learner’s capability, and it should be specific and clear enough so students have the opportunity to change their performance; otherwise, students can become frustrated (Shute, 2008). Below are some assessment approaches that can incorporate these principles: class discussions, differentiated instruction, response to intervention, data-driven instruction, self-assessment, and peer assessment.

Class discussions provide valuable opportunities to reveal students’ levels of understanding, reasoning processes, questions, and misconceptions, all of which help to inform instruction during the discussion and in future lessons. Engaging learners in dialogues that build on a series of questions helps students to develop their reasoning processes and a deeper understanding of a subject in real time (Wiliam, 2010). Effective discussions have questions with multiple or few known answers, ask students to provide support for interpretations, and have opportunities to discuss strategies related to interpretations. Teachers can ask questions like, “How do you know? What makes you think that?" and use techniques to elicit complex expression and understanding (for example, “Tell me more. What do you mean by ___?”) (Goldenberg, 1992/1993; Murphy, Wilkinson, Soter, Hennessey, & Alexander, 2009).

When feedback is given with details about the correct answer, students learn more than if they are just told whether the answer is correct or not (Wiliam, 2010). Discussions that support text comprehension also include a focus that connects with learners’ background knowledge (for example, using the text or personal experience) and require teachers and students to make comments that build on what others have said (Goldenberg, 1992/1993; Murphy et al., 2009). Some research suggests that the level of complexity in student responses rises the longer a teacher pauses after asking questions (Rowe, 1974). It is also crucial for students to deepen their knowledge by generating their own questions and for teachers to pay close attention to the questions that students ask.

Reciprocal teaching is a specific dialogue technique that may enhance reading comprehension through “guided practice in applying simple, concrete strategies to the task of text comprehension” (Rosenshine & Meister, 1994). The approach involves teachers modeling the strategies of questioning, clarifying, summarizing, and predicting. Teachers use dialogue to apply the strategies as students read the text and then gradually withdraw support as students become more competent. Students may shift into groups in which they ask and answer the questions with one another. Edutopia’s coverage of The College Preparatory School in Oakland, California, shows many effective discussion techniques in action.

Differentiated instruction involves tailoring the broad range of formats, activities, and assessments that can be used to learn and show mastery of a subject to students’ unique abilities, needs, and interests. Differentiated instruction often involves creating groups within a classroom, grade level, or school so learners can receive more-personalized instruction that addresses their learning needs.

For example, Edutopia’s video of the Reteach and Enrich program at Mesquite Elementary School in Arizona shows how a grade-level team uses collaborative lesson planning to differentiate instruction. According to weekly assessment results, students who need more support with specific concepts receive reteaching, while students who grasped the concepts participate in enrichment activities.

Differentiated instruction may also work with whole-school reforms. For example, in a random-assignment experiment, differentiated instruction was equally if not more effective in improving reading fluency when compared to the traditional whole-group approach (Reis, McCoach, Little, Muller, & Kaniskan, 2011). In this study, differentiated instruction began with a book discussion and read aloud, with time for independent reading, and integrated reading strategies or higher-level-thinking questions. Students then received differentiated reading strategies in five-minute individual conferences or participated in literary discussions, and groups had options for independent reading, creativity training, buddy reading, or other choices.

Response to Intervention (RtI) is a school-wide or district-wide approach that incorporates differentiated instruction and provides three levels, or tiers, of increasingly intense support based on students’ performances on benchmark assessments. At Tier 1, all students receive universal screening in reading and math to identify students who are “at risk” and provide interventions while regularly monitoring progress. Students who score below the universal screening benchmark receive Tier 2 interventions. In reading, small groups meet three to five times a week for 20 to 40 minutes and receive systematic instruction in up to three foundational reading skills. In math, small groups typically meet four to five times per week for 20 to 40 minutes and receive targeted instruction aimed at building math proficiencies (Gersten, Beckmann, Clarke, Foegen, Marsh, Star, & Witzel, 2009). Students who are not making sufficient progress in Tier 2 receive Tier 3 instruction, which focuses on a few skills in daily sessions that are often one-to-one.

RtI practices have been shown to support reading achievement at the elementary school level (Gersten, Compton, Connor, Dimino, Santoro, Linan-Thompson, & Tilly, 2008) and math achievement at the elementary school and middle school levels. Further information on schools and districts that have implemented RtI in reading and math can be found on the Doing What Works Public Files website.

Data-driven instruction involves collecting, interpreting, and disseminating data to improve instructional decision making in a coordinated district-wide and school-wide effort. Data about student learning, demographics, school processes, and teacher perceptions are used to inform decision making, and extensive professional development is used to set goals, prioritize, and make appropriate intervention plans (Slavin, Cheung, Holmes, Madden, & Chamberlain, 2012).

District-wide data-driven reform efforts have been shown to increase math and reading achievement (Carlson, Borman, & Robinson, 2011; Slavin et al., 2012). However, it is important to note that consultations or benchmark assessments did not in and of themselves change achievement; rather, it was the fact that consultations and benchmarks motivated schools to change teaching and learning through appropriate, proven, replicable strategies that resulted in improved achievement (Slavin et al., 2012).

Delivering frequent, personalized feedback to each student may seem like a daunting task. One way to help “free up” teachers’ time so they can provide more-individualized support is to diffuse responsibility for providing feedback through self-assessment and peer-assessment activities that activate learners as owners of their own learning and instructional resources for themselves and their peers (Black & Wiliam, 2009).

Self-assessment engages students in applying criteria to analyze and to improve their own work; however, it is not recommended for use as a final assessment, in which students determine their own grades (Andrade & Valtcheva, 2009). Self-assessment can improve student performance and help students become self-correcting learners when they understand the criteria and ways to apply them and when they receive feedback in using self-assessment data to improve their performance (Andrade & Valtcheva, 2009).

Peer assessment engages students in analyzing one another’s work against the same criteria they use to check their own learning and providing peers with feedback that helps them to progress toward the lesson goal. Peer assessment may improve performance for students who receive evaluations and for students as they give evaluations of writing assignments (Karegianes, Pascarella, & Pflaum, 1980) at the elementary school and secondary school levels (Topping, 2009). The effectiveness of peer feedback is improved with the training that student assessors receive, the clarity of evaluation criteria (Topping, 2009), and the extent to which student assessors provide justifications for their assessments (Gielen, Peeters, Dochy, Onghena, & Struyven, 2010).

The methods for gathering evidence of understanding and providing rapid feedback are as varied as the teachers who invent them. For further examples, see Grant Wiggins’s (coauthor of Understanding by Design and author of Educative Assessment) 13 practical examples and these helpful formative-assessment examples from Edutopia community-group contributor David Wees. Edutopia’s research review on Using Technology for Assessment and Customizing Instruction also provides technology-integrated assessment practices.

Continue to the next section of the Comprehensive Assessment Research Review, Motivating Students.