Assessing When Numbers Don`t Count

Assessing When Numbers
Don’t Count
Binghamton University
March 23, 2007
Today’s Objectives




Define what a discourse-based, or
qualitative assessment method is
Review a few of these methods
Work with some case studies to better
understand how these methods might be
used
Discuss the uses and limitations of these
methods, especially with regard to
assessment
The Pressure to Assess




This week—Spellings Commission meeting to
discuss several issues. One of them is student
learning outcomes
There is consistent pressure to use standardized
tests and surveys; we are currently using one of
them, the NSSE
Through the use of standardized tests and
surveys, we gain perspective on how we
compare to other institutions
We might be able to gain a “value added”
perspective
Weaknesses of the Standardized
Test/Survey Approach





We obtain a “macro” perspective, but may not
gain a “micro” perspective
We have little control over the questions
“Value added” is still difficult to establish with
standardized tests
Faculty see little value in using standardized
tests, especially in interdisciplinary studies and
the liberal arts
Standardized tests/surveys often assume they
can measure most of what concerns faculty in
respect to teaching and learning
Forging the Middle Ground:
Discourse-Based Assessment Methods






Allow for the discovery of the unanticipated
Relevant to interdisciplinary study
Maximizes faculty/staff input when discourse is
highly valued
Good to use when the number of objectives
outweigh the amount of time available to assess
student learning
Great contextualizer
When performed carefully and compared with
other methods, great return on investment
Types of Qualitative Assessment







Focus groups
Expert panels
Open-ended surveys
Ethnographic studies (participant
observations)
Portfolio reviews
Primary trait scoring
Delphi panels
Qualitative Assessment is an
Inductive Process
Defined Focus
Observations
Analysis
Summary Report
Comparison with Other
Assessments
Action
Expert Panel






Type of focus group
Focus is on (for assessment purposes) a
particular assignment or performance
Not a simple conversation; is done methodically,
with precision, and is systematic
Often involves convenience or random samples
of a homogeneous population
Must include carefully written questions
Might include a survey or other assessment
technique as part of the process
Krueger’s 10 Quality Factors in Focus
Group (Expert Panel) Research










Clarity of purpose
Appropriate environment
Sufficient resources
Appropriate participants
Skillful moderator
Effective questions
Careful data handling
Systematic and verifiable analysis
Appropriate presentation
Honoring the participant, clarity, and method
Expert Panel Procedure
Select Focus
Report Results
Select Experts
Write
Questions
Conduct Expert Panel—
How to Assess Agreement?
Use Additional
Method?
Logistics
Final Thoughts on Expert Panels





Good method of assessing “ineffable outcomes”
Better when expert panel questions and
conversations are grounded in standards and/or
student learning objectives
Also good when specific focus is on a particular
assignment or performance
It is advisable to use a secondary method either
prior to or during expert panel
Does not control for anonymity among
respondents
Primary Trait Scoring



Focus is upon one particular assignment,
performance, etc., that is reflective of
several aggregate student learning
outcomes
Rate each outcome according to a scale—
ex., proficient, satisfactory, unsatisfactory
Idea is to look at trends, not numbers,
that spark discussion
Primary Trait Scoring--Procedure






Choose an assignment in which students demonstrate
summative knowledge, skills, or competencies
Carefully rate student performance according to the
scale
Place checkmarks in each column
Look for visual trends
Discuss why these trends occur, what basis these rating
occurred, and what specific issues are revealed through
the analysis
Combine with other findings, or make plans for action
Example of Primary Trait Analysis
Element
Excellent
Satisfactory
Unsatisfactory
Demonstrate an ability
to research the
financial integrity
of a business plan
through
accounting/financ
ial analysis
√√√√
√√√√√ √√
√
Demonstrate an ability
to assess a plan’s
organizational
integrity; that is,
a business plan
demonstrates how
an organization
will be built and
sustained to
under gird the
success of the
business being
organized
√√
√√√√
√√√√√
√
Primary Trait Analysis—Final
Thoughts




Method to get faculty or staff to talk about
what assessment results mean
A good starting point toward developing a
rubric
Enables discussion, which can lead to
further discovery
“Simple, stress free, and easy”
Delphi Panel Introductory Exercise




Divide into 3 groups
Get out piece of paper and individually
write down, “what do students have the
most difficulty with when they first come
to college (as first-year students)?”
Try to create frequency counts—combine
like answers and tally them
Discuss
Questions for Groups



What do these say about the difficulty
students might have when they start?
Take a look at the most “popular”
answer—do these ordinarily achieve
“majority vote status?
Even in cases where “majority vote status”
is achieved, might less popular answers
indicate group consensus?
Introduction to the Delphi Method



Combination of at least 3 methods—open-ended
survey, closed-ended survey, and expert panel
Unlike expert panel, attempts to maximize
anonymity of respondents to control for power
dynamics among these respondents
Assumes highly motivated groups of experts
(faculty or staff) willing to participate in more
than one round of questions
Introduction to the Method
1.
2.
3.
4.
5.
6.
Find homogeneous group of experts who can comment
either on one assignment or specific student learning
outcomes
Create an open-ended survey in which respondents are
asked to identify strengths and weaknesses in student
performance in reference to specific standard or student
learning outcome
Content analyze responses by combining like responses,
placing how many times each was mentioned in
parentheses
“Cut and paste” these onto a survey, and ask respondents
to indicate to what extent they agree with each on a 4 or
5 point scale
Report those responses that indicate consensus
If needed, move to 3rd round, in which respondents rank
these consensus items
Strengths of Delphi Method



A way of addressing “ineffable outcomes”
Can be used to designate the most
agreed-upon student learning objectives
that faculty have communicated
Can be used to gather information from
employers, internship supervisors, alumni,
etc. about specific items of interest
Limitations of Delphi Method



Can be time consuming
Takes some knowledge of statistics
Not a method that can be used by itself;
usually results need to be compared with
direct assessments of student learning
Today’s Activities




Separate into three groups; select group
note taker
If you have not already, read the case
study packets
As a group, discuss questions at end of
case study—debate, applaud, etc.—do
something active
Write answers on provided sheet of paper