★ Must-know
📌 Grade retention keeps a student in the same grade, whereas social promotion advances the student to the next grade with their peers.
Research commonly associates retention with:
In a study of almost 12,000 kindergarten students followed through the end of first grade, retention showed no evidence of improving reading or mathematics achievement or making first-grade classes more similar in academic ability. - G. Hong and Raudenbush, 2005
📌 Students having academic difficulty should receive help whether they are promoted or retained; repeating the same material in the same way does not solve academic or social problems.
Further detail
In 2017, about 1,082,000 kindergarten-through-grade-12 students were retained in grade. - Digest of Educational Statistics, 2020
A Florida study found that students retained in third grade based on reading proficiency scores took fewer remedial high-school courses and were as likely to graduate as students who were not retained. - Schwerdt et al., 2015, 2017
About 4% to 5.5% of children delay kindergarten entry, and children who are red-shirted are most likely to be White, male, and from high socioeconomic-status families. - Bassok & Reardon, 2013
A study following retained students for four years found short-term social and behavioral advantages followed by long-term problems and vulnerabilities.
In a study following students retained in grades 1 through 5, retention increased school dropout, especially among Black and Hispanic girls. - Jan Hughes and her colleagues, 2018
A suggested alternative to retention is to promote students with their peers while providing special remediation during the summer or following year and support for attention and self-regulation skills.
Retention repeats a grade; social promotion advances students with their peers.
★ Must-know
Grades may support motivation when teachers grade in students' best interests and help students interpret grades as feedback for learning. - McMillan, 2019
Instead of failing work, teachers can mark it incomplete and support revision while maintaining high standards and giving students an opportunity to meet them.
Teachers should explain grading policies early, use clearly specified reasonable standards, and base grades on as much objective evidence as possible.
A number or letter grade does not convey the totality of a student's experience in a class, so communication with families should involve more than sending grades.
Further detail
Low grades can lead students to:
Low grades at the start of the year can create a downward spiral in which lower grades reduce engagement, which in turn contributes to lower grades.
Teachers should ensure students understand test directions, then correct, return, and discuss tests as soon as possible, including why popular wrong answers are incorrect.
Teachers should guard against grading bias, keep students informed of their standing, and generally change a grade only to correct a clerical or calculation error.
Because measurement involves error, teachers should give the benefit of the doubt in borderline cases and consider revising or discarding a question that many students miss in the same way.
Teachers should give credit for creative, partially correct, and rationally argued answers rather than reserving high grades and praise only for answers that match the teacher's or textbook's views.
To give students a reasonable chance to succeed, teachers can check prerequisite skills, offer equally difficult retests, treat early failing work as incomplete, and give greater weight to work at the end of a unit.
Meaningful grading balances written and oral feedback, ties grades to important learning targets, allows ungraded exploration, and uses more than one criterion, such as essays, multiple-choice items, oral reports, and participation.
Ways to communicate with families include (Guskey and Bailey, 2001):
📌 Family conferences should be friendly and unrushed, use factual observations, listen to concerns, and keep information from students or caregivers confidential.
Grades used as feedback can support motivation; low grades alone can trigger disengagement.
★ Must-know
Further detail
Very high or very low scores affect the mean, so the median may better represent central tendency when a group contains a few extreme scores.
In a normal distribution, about 16% of scores are more than one standard deviation above the mean, and about 2% are more than two standard deviations above it.
In 2020, the SAT Math section had a mean of 523 and a standard deviation of 117, so a score of 640 was one standard deviation above the mean. - College Board
Picture a bell with most scores clustered around its midpoint and fewer scores in its tails.
★ Must-know
📌 A percentile rank of 75 means that the student's score exceeds that of 75% of the norm group; it does not mean 75 correct answers or 75% of test questions correct.
Further detail
A percentile rank of 50 indicates a score better than 50% of the norm group and is described as average.
Percentile-rank gaps do not correspond to constant raw-score gaps: the 50th-to-60th percentile difference may be two raw points, while the 90th-to-99th percentile difference may be about ten.
📌 A high grade-equivalent score does not establish that a student is ready for advanced work, because different grade-level test forms may assess different material.
📌 Grade-equivalent units do not have the same meaning at every grade level, and educators and psychologists consider these scores misleading and often recommend not using them.
📐 Formula - The T score is calculated as .
Percentiles rank students; standard scores express distance from the mean.
★ Must-know
Achievement-test publishers commonly provide individual student profiles showing scores on each subtest.
A National PR-S such as 59-5 reports a percentile rank of 59 and a stanine of 5; percentile rank indicates the percentage of students at the same grade level with lower scores.
A National Grade Percentile Band estimates the range in which a student's true score is likely to fall by adding and subtracting the test's standard error from the observed score; nonoverlapping bands indicate likely achievement differences.
Further detail
A student report may include a Lexile Measure computed from the Reading Comprehension score to help teachers select texts appropriate to the student's reading level.
The sample report presents achievement results for:
The sample student report also includes total battery scores and scores on the Otis-Lennon School Ability test, a group IQ or scholastic aptitude test.
A National Normal Curve Equivalent is derived from percentile rank and ranges from 1 to 99, with a mean of 50 and a standard deviation of 21.
An Achievement/Ability Comparison range compares a student's subtest achievement with students of similar ability measured by the Otis-Lennon School Ability test and classifies the result as HIGH, MIDDLE, or LOW.
In a score report's skill clusters, NP is the number of questions possible, NA is the number attempted, and NC is the number correct; skills may be assessed with only 3 to 8 questions, and fewer items mean lower reliability.
★ Must-know
College academic success is associated only partly with SAT or ACT scores: 25% is associated with those scores, while 75% is associated with motivation, hard work, study habits, interests, and other factors under the student's control. - Popham, 2020
Effective family conferences should:
📌 When explaining test results, teachers should use nontechnical language, explain the comparison group for norm-referenced tests or the assessed tasks for criterion-referenced tests, and describe scores as ranges rather than perfect single-point measures.
Further detail
High school grades predict on-time college graduation better than admission-test scores partly because grades provide information about self-regulation and motivation. - Galla et al., 2019
For norm-referenced tests, percentile scores are easier to explain than other score types: 50 is average and 99 is the highest possible percentile, and a percentile is not the percentage of items answered correctly.
Teachers should avoid grade-equivalent scores when explaining results and clarify that a high grade equivalent reflects thorough understanding of the current grade, not capacity for higher-grade work, and that the same score can mean different things across subjects.
★ Must-know
Decisions based on test scores require distinguishing test quality from test use and considering who is tested, the consequences and effects of testing, interpretation across student groups, what constructs are measured, and how scores are combined with other information.
Testing should not serve accountability alone, because using it only to estimate yearly learning can displace instructionally relevant feedback for teachers and students.
High-stakes results should not be used for pass/fail decisions without strong evidence that the test is valid, reliable, and free of bias, and a single test should not determine pass/fail outcomes. - Haladyna, Essentials of standardized achievement testing: Validity & accountability, 2002
Further detail
Test results can affect:
In some states, about 80% of elementary schools spend about 20% of instructional time preparing for end-of-grade tests. - Abrams & Madaus, 2003
High-stakes testing can narrow curricula because tested material receives attention while untested material is neglected, and reading demands in mathematics tests can disadvantage students with weak reading skills or limited English proficiency. - Haertel, 2018
In a study of 8 million fourth- and eighth-grade students in 47 states, item format explained about 25% of variation in state gender gaps: males generally did better on multiple-choice items and females on essay items. - Reardon et al., 2018
📌 State-to-state test-score comparisons are not meaningful because states differ in curricula, tests, resources, and challenges.
📌 Test scores alone are inappropriate for evaluating teachers or schools because family and community resources, student mobility, and other factors may affect scores beyond educators' control.
A test can inform instruction, but when accountability becomes its sole purpose, feedback gets lost.
★ Must-know
📌 A sound high-stakes testing program uses reliable, valid, unbiased tests aligned with district standards, embeds them in a broader assessment plan, assesses complex thinking, provides alternate assessments and retesting when appropriate, includes all students with informative reporting, offers remediation, ensures learning opportunities, accounts for students' first language, and uses results for rather than against children. - Haladyna, Essentials of standardized achievement testing: Validity & accountability, 2002
Further detail
PARCC and SBAC were state consortia funded by the federal government in the early 2010s to develop assessments aligned with the Common Core Standards and intended to measure complex thinking through varied approaches, including performance assessment.
A preparation program for students with disabilities taught test-taking strategies over six class periods, including completing answer-sheet bubbles, sorting problems by difficulty, estimating with rounding, underlining key question words, and eliminating alternatives with redundant information or extreme qualifiers. - Carter et al., 2005
After the preparation program, students with disabilities improved their test scores significantly, but most did not reach the passing level; the authors recommended beginning preparation earlier. - Carter et al., 2005
📌 Teachers preparing students for tests should align test questions with learning goals, check coverage and testing barriers, familiarize students with materials and formats, follow administration instructions and time limits exactly, and create a calm, quiet environment while monitoring for cheating.
📌 Students should prepare by studying and sleeping well, arriving ready to concentrate, reading directions and questions carefully, managing time by answering easier questions first, guessing according to scoring rules, checking answers, answering essays directly, and learning from test feedback.
★ Must-know
📌 In value-added modeling, achievement above predicted growth yields a positive estimated teacher or school effect, achievement at the predicted level yields zero, and achievement below prediction yields a negative effect.
📌 Value-added scores should be based on several years of data from many students and used to identify school or curriculum strengths and weaknesses and guide professional development, not to evaluate individual teachers. - American Educational Research Association Council, 2015
Further detail
📌 Value-added judgments require valid, reliable, curriculum-aligned tests with score ranges broad enough to capture achievement above and below grade level, and teachers should have students for most of the year.
Value-added estimates become more uncertain when focused on a single class rather than a whole school, and standard tests may inadequately measure growth among students far above or below grade level.
Comparing prior-spring scores with current-year scores can include summer learning loss in the measured change even though the teacher did not cause that loss. - Atteberry & Mangan, 2020
Teacher effects cannot be cleanly isolated when students and teachers are not randomly assigned to classes, because student selection and conditions such as materials, support, tutoring, and time for homework can affect scores. - Popham, 2020
📌 Kimberlee Everson argued that teacher evaluation requires answering fundamental questions about whether good teaching means achieving academic goals alone or nonacademic goals as well, and whether all students should reach a set standard regardless of background.
If actual growth exceeds predicted growth, the estimated teacher or school effect is positive.
★ Must-know
In a study of about 21,000 New Mexico teachers evaluated from 2015 to 2016, 40% would have received a different composite score if reevaluated in the same year, according to Sy Doan and colleagues (2019).
Evaluating teachers requires deciding whether good teaching means reaching academic goals, achieving nonacademic goals too, bringing all students to a set standard, or helping each student progress as far as possible.
Further detail
In the New Mexico evaluation study, estimates were less consistent at reevaluation when value-added information had greater weight in composite scores.
The validity of teacher evaluation depends on whether the evaluation method matches the conception of good teaching being used.
VAM estimates score gains; teacher evaluation must also establish what counts as good teaching.
★ Must-know
The MET Project identified three measures to use together:
Classroom observation ratings can vary with the students being taught, and class composition may bias ratings; male teachers and teachers with high concentrations of Black, Latino, male, and low-performing students receive significantly lower ratings, according to Campbell and Ronfeldt (2018). - Campbell & Ronfeldt
Further detail
The Bill and Melinda Gates Foundation launched the Measures of Effective Teaching (MET) Project in 2009 as a research partnership involving 3,000 teachers and research teams at dozens of institutions.
The Tripod Student Perception Survey was developed by Ron Ferguson at Harvard University and asks students to respond to statements about their teachers.
Tripod survey statements are tailored to age groups, including remembering learning for K–2 students, correcting mistakes for upper elementary students, and giving full effort for secondary students.
Student Tripod scores and observer scores from the Danielson Framework predict student academic growth on achievement tests, according to Sandilos et al. (2019).
Take a 37-question practice quiz on Assessment, Grading, and Test Scores, with every answer explained.
1. A school keeps Maya in the same grade for another year, while her classmates advance. Which approach is being used for Maya?
2. Which set of outcomes is commonly associated with grade retention in research?
Memorize the key concepts of Assessment, Grading, and Test Scores with 87 flashcards.
What does grade retention do to a student's grade placement?
It keeps the student in the same grade.
What outcomes is grade retention commonly associated with?
Dropping out, higher arrest rates, fewer job opportunities, and lower self-esteem.
What did G. Hong and Raudenbush, 2005 find about retention's academic effects?
Retention showed no evidence of improving reading or mathematics achievement.
Upload your class notes and Revizly writes a study guide, a quiz and flashcards in about a minute.
Study guide maker