Some principles of assessment audits - PowerPoint PPT Presentation

1 / 16
About This Presentation
Title:

Some principles of assessment audits

Description:

Score / results interpretation guides appropriate for each category of user, e.g. ... Provide score / results recipients with an appropriate frame of reference for ... – PowerPoint PPT presentation

Number of Views:28
Avg rating:3.0/5.0
Slides: 17
Provided by: Clifford66
Category:

less

Transcript and Presenter's Notes

Title: Some principles of assessment audits


1
Some principles of assessment audits
  • Cliff Adelman, Institute for Higher Education
    Policy, Feb. 29, 2008

2
Whats this about?
  • Documentation
  • Design
  • Communication
  • Validities
  • Data Mapping
  • Accountability

3
If you are designing an assessment program, you
will be either
  • Using off-the-shelf products
  • Modifying or reweighting available products
  • Developing your own products / procedures
  • Drawing on unobtrusive data information

4
The integrity of what you do
  • requires an vigilant audit structure and process.

5
Documentation right from the start
  • Record keeping of everything you did
  • A single location of records, e.g. IR office
  • Scheduled review of records
  • Assessment data library as part of records
  • Why? Whether the subject of assessment is the
    student, the program, or the institution, youre
    on sensitive territory. Documentation is a form
    of protective anticipation.

6
Audit components design
  • What do you want to know?
  • About who or what?
  • Process options in light of answers to those
    questions.
  • Virtues / limitations of each option considered.

7
Audit components content before populations
  • Critical content, critical skills
  • Weighting of domains says who?
  • Rationale documentation
  • Level(s) of difficulty / challenge says who?
  • Distribution of difficulty / challenge
  • Sufficiency of prompts

8
Audit components populations
  • Full census? why?
  • Samples? of who?
  • If sample, sufficiency of subgroups
  • Nature and adequacy of population background
    information
  • Time and population

9
Relationship of instrument / method to judgment
  • Conditions for ensuring maximum reliability
  • Differences between restricted / unrestricted
    response
  • Conditions for performance assessments
  • Training for reliability (unrestricted and
    performance)

10
Directions and presentation
  • Visual form fonts, screens, layout
  • Clarity of directions in relation to the type of
    assessment
  • Completeness of directions
  • All of this pre-tested with representative focus
    groups
  • Document revisions!

11
Indeed, pre-testing of everything
  • With records of administration, time,
    reliability, item-analysis (where appropriate),
    etc.

12
And if its a test, documented reviews of. . .
  • Adequacy of fit of item-response models to data
  • Test characteristics compared with its
    psychometric specifications
  • Test editions, to ensure content
    representativeness
  • Test forms, to assure comparability of scoring

13
Prior information for prospective examinees, i.e.
Do they know. . .?
  • Purpose of the assessment
  • What the assessment will be like, through samples
    of prompts
  • Experience relevant to assessment performance
  • Administration procedures
  • The best documentation is a sample information
    bulletin.

14
Reliability of any assessment
  • Document sources of variation (content, judges,
    time interval between assessment and judgment,
    etc.)
  • Document methods used to determine reliability
    the rationale for using them
  • Document results of reliability analysis, e.g.
    reliability coefficient, standard error of
    measurement, degree of agreement between
    independent judgments

15
Post-Assessment Communication
  • Score / results interpretation guides appropriate
    for each category of user, e.g. instructor,
    institution, agency, legislature, media
  • Recommend only those interpretations for which
    supporting information is available
  • Strong statement of valid uses of results
    equally strong statement of invalid uses of
    results
  • Provide score / results recipients with an
    appropriate frame of reference for evaluating the
    performance represented in the scores / results
  • Be consistent!

16
We could cover a lot more
  • But this is a reasonable beginning on the process
    and content of assessment audits
  • Much of it derives from the internal technical
    manuals of ETS and ACT
  • You certainly could elaborate based on your
    experience, but
  • Always think like an auditor!
Write a Comment
User Comments (0)
About PowerShow.com