Assessment Primer: How to do Assessment

Assessment to improve learning focuses on how well students are learning what we intend them to learn.  By establishing learning objectives, assessment methods are used to measure selected learning outcomes to see whether or not the objectives have been met for the course or program.

How Cycle


Ways of gathering evidence

  • Formative – those undertaken while student learning is taking place; the purpose or which is to improve teaching and learning; designed to capture students’ progress toward institution- or program-level outcomes based on criteria and standards of judgment
  • Summative – those obtained at the end of a course or program; the purpose of which is to document student learning for transcripts and for employers, donors, legislators, and other external audiences; designed to capture students’ achievement at the end of their program of study and their undergraduate or graduate education based on criteria and standards of judgment
  • Direct – evidence of student learning which is tangible, visible, self-explanatory; prompt students to represent or demonstrate their learning or produce work so that observers can assess how well students’ texts or responses fit institution- or program-level expectations
    • Example: performances, creations, results of research or exploration, interactions within group problem solving, or responses to questions or prompts
  • Indirect – evidence which provides signs that students are probably learning, but the evidence of exactly what they are learning is less clear and less convincing; capture students’ perceptions of their learning and the educational environment that supports that learning, such as access to and the quality of services, programs, or educational offerings that support their learning
    • Example: student satisfaction, alumni, and employer surveys
  • Objective – one that needs no professional judgment to score correctly (although interpretation of the scores requires professional judgment); examples: multiple-choice, true-false exams
  • Subjective – yield many possible answers of varying quality and require professional judgment to score
  • Traditional – the kinds of tests that have been around for decades; e.g., objective tests, ‘blue book’ essay questions, and oral examinations
  • Performance – ask students to demonstrate their skills rather than relate what they have learned through traditional tests; e.g., field experiences, laboratory and studio assignments, projects.  Also called authentic assessments when asking students to do a real-life task.  Have two components: (i) the assignment or prompt that tells students what is expected of them and (ii) a scoring guide or rubric used to evaluate completed work.
  • Embedded – program assessments which are embedded into course work
  • Add-on – assessments which are in addition to course requirements; e.g., assemble a portfolio, take a standardized test, participate in a survey
  • Local – created by faculty and/or staff
  • Published – those published by an organization external to the institution and used by several institutions
  • Quantitative – use structured, predetermined response options that can be summarized into meaningful numbers and analyzed statistically; place interpretative value on numbers; e.g., the number of right versus wrong answers
  • Qualitative – use flexible, naturalistic methods and are usually analyzed by looking for recurring patterns and themes; e.g., reflective writing, notes from interviews and focus groups; place interpretative value on the observer; e.g., observations of group interaction or an individual’s performance in a simulation

(Assessing Student Learning: A common sense guide by Suskie 2004 and Assessing for Learning: Building a sustainable commitment across the institution by Maki 2004)

A summary of Direct Assessment Techniques

A summary of Indirect Assessment Techniques

A summary of Survey Formats

A summary of the Pros and Cons of Assessment Tools

Choosing the Right Assessment Tool


Using Rubrics for Direct Assessment of Student Work

What is a rubric?

A rubric is a scoring tool that lays out the specific expectations for an assignment.  Rubrics divide an assignment into its component parts and provide a detailed description of what constitutes acceptable or unacceptable levels of performance for each of those parts.

(Introduction to Rubrics: An Assessment Tool to Save Grading Time, Convey Effective Feedback, and Promote Student Learning by Stevens and Levi 2005)

What are the parts of a rubric? Rubrics are composed of four basic parts:

  • A task description (the assignment)
  • A scale of some sort (levels of achievement, possibly in the form of grades).  Scales typically range from 3 to 5 levels.
  • The dimensions of the assignment (a breakdown of the skills/knowledge involved in the assignment)
  • Descriptions of what constitutes each level of performance (specific feedback)


  • Can be used to classify virtually any product or behavior, such as essays, research reports, portfolios, works of art, recitals, oral presentations, performances, and group activities
  • Can be used to provide formative feedback to students, to grade students, and to assess programs
  • Can be used for program assessment in a number of ways:
  • Faculty can use rubrics in classes and aggregate the data across sections
  • Faculty can independently assess student products and then aggregate results
  • Faculty can participate in group readings in which they review student products together and discuss what they have found

(Assessing Academic Programs in Higher Education by Allen 2004)

How to Create a Rubric


Rubric Formats

Rubrics can be classified into four formats
(Assessing Student Learning: A common sense guide by Suskie 2004)

  • Checklists – simple list indicating the presence of ‘things you are looking for’
    A checklist rubric for evaluating a web site
    (Assessing Student Learning: A common sense guide by Suskie 2004)
    X Titles are meaningful
    Each page loads quickly
    X The text is easy to read
  • Rating scales – a checklist with a rating scale added to show the degree to which the ‘things you are looking for’ are present
A rating scale rubric for an information literacy assignment
(Assessing Student Learning: A common sense guide by Suskie 2004)
Please indicate the student’s skill in each of the following respects, as evidenced by this assignment, by checking the appropriate box.  If this assignment is not intended to elicit a particular skill, please check the N/A box.
Ver y
Identify, locate, and access sources of information
Critically evaluate information, including its legitimacy, validity, and appropriateness
Organize information to present a sound central idea supported by relevant material in a logical order
Use information to answer questions and/or solve problems
Clearly articulate information and ideas
Use information technologies to communicate, manage, and process information
Use information technologies to solve problems
Use the work of others accurately and ethically
What grade are you awarding this assignment?
If you had to assign a final course grade for this student today, what would it be?


A rating scale for an oral presentation
(Assessing Student Learning: A common sense guide by Suskie 2004)

The presenter …
Strongly agree Agree Disagree Strongly disagree
Clearly stated the purpose of the presentation X
Was well organized X
Answered questions authoritatively X
Appeared confident X


It should be noted that rating scales can be vague in nature leading to problems (Suskie 2004):

  • When several faculty are doing the rating, they may be inconsistent in how they rate performance
  • Students don’t receive thorough feedback; i.e., a scored rubric may not explain why something was less than superior
  • Holistic rating scales
    • Do not have a list of the ‘things you’re looking for’
    • Have short narrative descriptions of the characteristics of outstanding work, acceptable work, unacceptable work, and so on
      HOLISTIC rubric for assessing student essays
      (Assessing Academic Programs in Higher Education by Allen 2004)
      Inadequate The essay has at least one serious weakness.  It may be unfocused, underdeveloped, or rambling.  Problems with the use of language seriously interfere with the reader’s ability to understand what is being communicated.
      Developing competence The essay may be somewhat unfocused, underdeveloped, or rambling, but it does have some coherence.  Problems with the use of language occasionally interfere with the reader’s ability to understand what is being communicated.
      Acceptable The essay is generally focused and contains some development of ideas, but the discussion may be simplistic or repetitive.  The language lacks syntactic complexity and may contain occasional grammatical errors, but the reader is able to understand what is being communicated.
      Sophisticated The essay is focused and clearly organized, and it shows depth of development.  The language is precise and shows syntactic variety, and ideas are clearly communicated to the reader.
  • Descriptive rubrics
    • Replace the checkboxes of rating scale rubrics with brief descriptions of the performance that merits each possible rating
    • Descriptions of each performance level make faculty expectations explicit and student performance convincingly documented.  But, coming up with succinct but explicit descriptions of every performance level for every ‘thing you are looking for’ can be time-consuming.
    • Are a good choice when several faculty are collectively assessing student work, it is important to give students detailed feedback, or outside audiences will be examining the rubric scores.

    A descriptive rubric for a slide presentation on findings from research sources
    (Assessing Student Learning: A common sense guide by Suskie 2004)

    Well done
    Needs improvement
    Organization Clearly, concisely written.  Logical, intuitive progression of ideas and supporting information.  Clear and direct cues to all information. Logical progression of ideas and supporting information.  Most cues to information are clear and direct. Vague in conveying viewpoint and purpose.  Some logical progression of ideas and supporting information but cues are confusing or flawed. Lacks a clear point of view and logical sequence of information.  Cues to information are not evident.
    Introduction Presents overall topic.  Draws in audience with compelling questions or by relating audience’s interests or goals. Clear, coherent, and related to topic. Some structure but does not create a sense of what follows.  May be overly detailed or incomplete.  Somewhat appealing. Does not orient audience to what will follow.


    ANALYTIC rubric for peer assessment of team project members
    (Assessing Academic Programs in Higher Education by Allen 2004)
    Below expectation Good Exceptional
    Project contributions Made few substantive contributions to the team’s final product Contributed a “fair share” of substance to the team’s final product Contributed considerable substance to the team’s final product
    Leadership Rarely or never exercised leadership Accepted a “fair share” of leadership responsibilities Routinely provided excellent leadership
    Collaboration Undermined group discussions or often failed to participate Respected others’ opinions and contributed to the group’s discussion Respected others’ opinions and made major contributions to the group’s discussion


ANALYTIC rubric for grading oral presentations
(Assessing Academic Programs in Higher Education by Allen 2004)
Below expectation Satisfactory Exemplary Score
Organization No apparent organization. Evidence is not used to support assertions. The presentation has a focus and provides some evidence that supports conclusions. The presentation is carefully organized and provides convincing evidence to support conclusions
(0 – 2) (3 – 5) (6 – 8)
Content The content is inaccurate or overly general.  Listeners are unlikely to learn anything or may be misled. The content is generally accurate, but incomplete.  Listeners may learn some isolated facts, but they are unlikely to gain new insights about the topic. The content is accurate and complete.  Listeners are likely to gain new insights about the topic.
(0 – 2) (5 – 7) (10 – 13)
Style The speaker appears anxious and uncomfortable, and reads notes, rather than speaks.  Listeners are largely ignored. The speaker is generally relaxed and comfortable, but too often relies on notes.  Listeners are sometimes ignored or misunderstood. The speaker is relaxed and comfortable, speaks without undue reliance on notes, and interacts effectively with listeners.
(0 – 2) (3 – 6) (7 – 9)
Total Score

Generic rubric for assessing portfolios
(Assessing Academic Programs in Higher Education by Allen 2004)

Evidence that the student has mastered this objective is not provided, unconvincing, or very incomplete
Evidence that the student has mastered this objective is provided, but it is weak or incomplete
Acceptable: Evidence shows that the student has generally attained this objective Exceptional:
Evidence demonstrates that the student has mastered this objective at a high level
Learning objective 1
Learning objective 2

Why use Rubrics?

  • Rubrics provide timely feedback – grading can be done more quickly
    Since students often make similar mistakes on assignments, incorporating predictable notes into the “descriptions of dimensions” portion of a rubric can simplify grading into circling or checking off all comments that apply to each specific student.
  • Rubrics prepare students to use detailed feedback
    In the rubric, the highest level descriptions of the dimensions are the highest level of achievement possible, whereas the remaining levels, circled or checked off, are typed versions of the notes/comments an instructor regularly writes on student work explaining how and where the student failed to meet that highest level.  Thus, in using a rubric the student obtains details on how and where the assignment did or did not achieve its goal, and even suggestions (in the form of the higher level descriptions) as to how it might have been done better.
  • Rubrics encourage critical thinking
    Because of the rubric format, students may notice for themselves the patterns of recurring problems or ongoing improvement in their work.
  • Rubrics facilitate communication with others
    TAs, counselors/tutors, colleagues, etc. can benefit from the information contained in the rubric; i.e., it provides information to help all involved in a student’s learning process.
  • Rubrics help faculty refine their teaching skills
    Rubrics showing a student’s continuing improvement or weaknesses over time, or rubrics showing student development over time, can provide a clearer view of teaching blind spots, omissions, and strengths.
  • Rubrics help level the playing field
    To aid first-generation or non-native speakers of English, rubrics can act as a translation device to help students understand what teachers are talking about.

(Introduction to Rubrics: An Assessment Tool to Save Grading Time, Convey Effective Feedback, and Promote Student Learning by Stevens and Levi 2005)

How can Rubrics be used to assess program learning goals?

  • Embedded course assignments – program assessments which are embedded into course assignments can be scored using a rubric
  • Capstone experiences – theses, oral defenses, exhibitions, presentations, etc. – can be scored using a rubric to provide evidence of the overall effectiveness of a program
  • Field experiences – internships, practicum, etc.—supervisor’s ratings of the student’s performance can be evidence of the overall success of a program
  • Employer feedback – feedback from the employers of alumni can provide information on how well a program is achieving its learning goals
  • Student self-assessments – indirect measures of student learning
  • Peer evaluations – while having the potential for being inaccurate and biased – they can motivate students to participate fully
  • Portfolios – rubrics can be a useful way to evaluate portfolios

(Assessing Student Learning: A common sense guide by Suskie 2004)

Rubric scores are subjective and thus prone to unintentional scoring errors and biases:

  • Leniency errors – when faculty judge student work better than most of their colleagues would judge it
  • Generosity errors – when faculty tend to use only the high end of the rating scale
  • Severity errors – when faculty tend to use only the low end of the rating scale
  • Central tendency errors – when faculty tend to use only the middle of the rating scale
  • Halo effect bias – when faculty let their general impression of a student influence their scores
  • Contamination effect bias – when faculty let irrelevant student characteristics (e.g., handwriting or ethnic background) influence their scores
  • Similar-to-me effect bias – when faculty give higher scores to those students whom they see as similar to themselves
  • First-impression effect bias – when faculty’s early opinions distort their overall judgment
  • Contrast effect bias – when faculty compare a student against other students instead of established standards
  • Rater drift – when faculty unintentionally redefine scoring criteria over time

(Assessing Student Learning: A common sense guide by Suskie 2004)

A Sample Rubric – The Critical Thinking Rubric

Example of a Rubric and its Use for Improvement

Reliability and Validity of Methods Used to Gather Evidence

Reliability and Validity narrow down the pool of possible summative and formative methods:

  • Reliability – refers to the extent to which trial tests of a method with representative student populations fairly and consistently assess the expected traits or dimensions of student learning within the construct of that method.
  • Validity – refers to the extent to which a method prompts students to represent the dimensions of learning desired.  A valid method enables direct and accurate assessment of the learning described in outcome statements.

(Assessing for Learning: Building a sustainable commitment across the institution by Maki 2004)


Reliable measures can be counted on to produce consistent responses over time:

  • Reliable data – variance in scores is attributable to actual differences in what is being measured, such as knowledge, performance or attitudes
  • Unreliable data – score variance is due to measurement error; which can include such things as the individuals responding to the instrument, the administration and scoring of the instrument, and the instrument itself

(Assessment Essentials: planning, implementing, and improving assessment in higher education by Palomba and Banta 1999)

Sources of error in regard to reliability:

(i) from the student

  • lack of motivation to take assessment seriously
  • prior experience with being evaluated
  • test anxiety, coaching, and physiological variables
(ii) from the assessment instrument

  • test items ambiguously worded
  • rating scale is confusing or vague
  • representativeness of test items; e.g., if two test forms differ in their emphasis on program content, inconsistent scores may result
(iii) from the conditions of assessment administration

  • varying the style of test administration procedures produces varying scores

The greater the error in any assessment information, the less reliable it is, and the less likely it is to be useful.
(Assessing Student Learning and Development: A Guide to the Principles, Goals, and Methods of Determining College Outcomes by Erwin 1991)

Types of reliability :

(i) Stability – usually described in a test manual as test-retest reliability; if the same test is readministered to the same students within a short period of time, their scores should be highly similar, or stable
(ii) Equivalence – the degree of similarity of results among alternate forms of the same test; tests should have high levels of equivalence if different forms are offere
(iii) Homogeneity or internal consistency – the interrelatedness of the test items used to measure a given dimension of learning and development
(iv) Interrater reliability – the consistency with which raters evaluate a single performance of a given group of students

(Assessing Student Learning and Development: A Guide to the Principles, Goals, and Methods of Determining College Outcomes by Erwin 1991)

Barriers to establishing reliability include rater bias – the tendency to rate individuals or objects in an idiosyncratic way:

  • central tendency – error in which an individual rates people or objects by using the middle of the scale
  • leniency – error in which an individual rates people or objects by using the positive end of the scale
  • severity – error in which an individual rates people or objects by using the negative end of the scale
  • halo error – when a rater’s evaluation on one dimension of a scale (such as work quality) is influenced by his or her perceptions from another dimension (such as punctuality)

Major Types of Reliability
(Assessing Academic Programs in Higher Education by Allen 2004)
Test-retest reliability A reliability estimate based on assessing a group of people twice and correlating the two scores.  This coefficient measures score stability.
 Parallel forms reliability (or alternate forms reliability) A reliability estimate based on correlating scores collected using two versions of the procedure.  This coefficient indicates score consistency across the alternative versions.
Inter-rater reliability How well two or more raters agree when decisions are based on subjective judgments.
Internal consistency reliability A reliability estimate based on how highly parts of a test correlate with each other.
Coefficient alpha An internal consistency reliability estimate based on correlations among all items on a test.
Split-half reliability An internal consistency reliability estimate based on correlating two scores, each calculated on half of a test.


Valid measures on ones in which the instrument measures what we want it to measure:

  • Construct-related validity – refers to the congruence between the meaning of the underlying construct and the items on the test or survey; i.e., do results correlate with other instruments examining the same construct?
  • Criterion-related validity – includes predictive validity: how dependable is the relationship between the scores or answers on an instrument and a particular future outcome?
  • Content-related validity – refers to the match between the content of the instrument and the content of the curriculum or other domain of interest

(Assessment Essentials: planning, implementing, and improving assessment in higher education by Palomba and Banta 1999)

Validity must be judged according to the application of each use of the method.  The validity of an assessment method is never proved absolutely; it can only be supported by an accumulation of evidence from several categories.  For any assessment methods to be used in decision making, the following categories should be considered:

  • Content relevance and representativeness
    • The selected test should be a representative sample from those educational objectives which the test is supposed to measure
    • The test should cover what the program covered and should place emphasis in proportion to the program’s emphases
    • Tests may be reliable but not valid for a particular program
  • Internal test structure
    • Typically demonstrated through intercorrelations among items covering the same content domain
  • External test structure
    • Necessary when the educator wishes to compare test scores or ratings with other measures or related variables
  • Process of probing responses
    • Typically sought at two points during any test or scale construction: initially in the test construction to determine whether the students’ interpretations are consistent with the intent of the test designer; and at the point of probing the process to see if a pattern might be discovered on those students who scored very high or very low
  • Test’s similarities and differences over time and across groups and settings
    • In studying validity evidence over time, some outcome measures should increase over time
  • Value implications and social consequences
    • If a test or rating scale discriminates against certain groups of people, that test or scale should be considered suspect.

(Assessing Student Learning and Development: A Guide to the Principles, Goals, and Methods of Determining College Outcomes by Erwin 1991)


  • involves establishing that an assessment measures what it is supposed to measure
  • can be thought of as the extent of the relationship between an assessment and the construct the assessment is supposed to predict
Major Types of Validity
(Assessing Academic Programs in Higher Education by Allen 2004)
Construct validity Construct validity is examined by testing predictions based on the theory (or construct) underlying the procedure.  For example, faculty might predict that scores on a test that assesses knowledge of anthropological terms will increase as anthropology students progress in their major.  We have more confidence in the test’s construct validity if predictions are empirically supported.
Criterion-related validity Criterion-related validity indicates how well results predict a phenomenon of interest, and it is based on correlating assessment results with this criterion.  For example, scores on an admissions test can be correlated with college GPA to demonstrate criterion-related validity.
Face validity Face validity is assessed by subjective evaluation of the measurement procedure.  This evaluation may be made by test takers or by experts in what is being assessed.
Formative validity Formative validity is how well an assessment procedure provides information that is useful for improving what is being assessed.
Sampling validity Sampling validity is how well the procedure’s components, such as test items, reflect the full range of what is being assessed.  For example, a valid test of content mastery should assess information across the entire content area, not just isolated segments.