OLF1206 101 Thalheimer

21
Beyond Smile Sheets: Learning Audits and Evaluations We Can Trust Will Thalheimer, Work-Learning Research Produced by December 7 & 8, 2006 101 Designing & Developing Online Assessments & Evaluations

Transcript of OLF1206 101 Thalheimer

Beyond Smile Sheets: Learning Audits and

Evaluations We Can Trust Will Thalheimer, Work-Learning Research

Produced by

December 7 & 8, 2006

101

Designing & Developing Online Assessments & Evaluations

Designing & Developing Online Assessments and Evaluations

December 7 & 8, 2006

Page 1Session 101 – Beyond Smile Sheets: Learning Audits and Evaluations We Can Trust, Will Thalheimer, Work-Learning Research

1

© Copyright 2006 Work-Learning Research, Inc.

Beyond Smile Sheets:

Will Thalheimer, PhDWork-Learning Research, Inc.

www.work-learning.comwww.willatworklearning.com

www.learningaudit.com

eLearning GuildDecember 2006

Some Research-Based Considerations

2

© Copyright 2006 Work-Learning Research, Inc.

Work-Learning Research, Inc.

MissionHelp Clients Build More Effective Learning Bridge Gap between Research & PracticeCompile Research, Share Findings

WorkLearning AuditsWorkshopsResearch

Designing & Developing Online Assessments and Evaluations

December 7 & 8, 2006

Page 2Session 101 – Beyond Smile Sheets: Learning Audits and Evaluations We Can Trust, Will Thalheimer, Work-Learning Research

Smile Sheet Example

4

© Copyright 2006 Work-Learning Research, Inc.

Four Levels (from Donald Kirkpatrick)

1. Reactionsmilesheets

2. Learningretrieval,

decision-making

3. Behavior on-the-jobperformance

4. Results businessresults

Designing & Developing Online Assessments and Evaluations

December 7 & 8, 2006

Page 3Session 101 – Beyond Smile Sheets: Learning Audits and Evaluations We Can Trust, Will Thalheimer, Work-Learning Research

5

© Copyright 2006 Work-Learning Research, Inc.

Percent of Companies Using

0 20 40 60 80 100

Reaction

Learning

Behavior

Results

1.

2.

3.

4.

79.5 %

35.0 %

15.3 %

8.8 %

ASTD’s industry reports

6

© Copyright 2006 Work-Learning Research, Inc.

For your most recent learning assessment, where did you give the learners the assessment?

A. In the training room/context.

B. In a different room/context.

C. Never done / Can’t remember.

Designing & Developing Online Assessments and Evaluations

December 7 & 8, 2006

Page 4Session 101 – Beyond Smile Sheets: Learning Audits and Evaluations We Can Trust, Will Thalheimer, Work-Learning Research

7

© Copyright 2006 Work-Learning Research, Inc.

For your most recent learning assessment, when did you give the learners the assessment?

A. At the end of training.

B. After a delay.

C. At end—and after a delay.

D. Never done / Can’t remember.

What do you What do you think aboutthink about

SMILE SHEETS?SMILE SHEETS?Type Your Answer in Chat

What do you think about Smile Sheets?

Designing & Developing Online Assessments and Evaluations

December 7 & 8, 2006

Page 5Session 101 – Beyond Smile Sheets: Learning Audits and Evaluations We Can Trust, Will Thalheimer, Work-Learning Research

9

© Copyright 2006 Work-Learning Research, Inc.

If smile-sheet ratings are high, what does that mean?

A. Learning results will likely be very good.

B. Learning results will likely be at least reasonably good.

C. Learning results cannot be predicted by smile sheets.

10

© Copyright 2006 Work-Learning Research, Inc.

Correlation between levels?

Level 2to

Level 3

r=.12

Level 1to

Level 3

r=.16

Level 1to

Level 2

r=.09

Alliger, Tannenbaum, Bennett, Traver, & Shotland (1997). A meta-analysis of the relations among training criteria.

Personnel Psychology, 50, 341-357.

Very Weak Relationship between Levels

Designing & Developing Online Assessments and Evaluations

December 7 & 8, 2006

Page 6Session 101 – Beyond Smile Sheets: Learning Audits and Evaluations We Can Trust, Will Thalheimer, Work-Learning Research

11

© Copyright 2006 Work-Learning Research, Inc.

“Did You Like Training” vs.

“Is the Training Valuable?”

Level 1to

Level 3

r=.07

Level 1to

Level 2

r=.02

Level 1to

Level 2

r=.15

Level 1to

Level 3

r=.20

Asking how much learners like training

Asking how valuabletraining was

BeyondBeyondSmile SheetsSmile Sheets

Beyond Smile Sheets

Designing & Developing Online Assessments and Evaluations

December 7 & 8, 2006

Page 7Session 101 – Beyond Smile Sheets: Learning Audits and Evaluations We Can Trust, Will Thalheimer, Work-Learning Research

13

© Copyright 2006 Work-Learning Research, Inc.

How We Learn

fosse

ditchRehearsal

“A fosse is a ditch.”

Long-Term Memory

14

© Copyright 2006 Work-Learning Research, Inc.

How We Retrieve Information

fosse

ditchSearch

Cue

Rehearsal

“A fosse is a ditch.”

Long-Term Memory

Designing & Developing Online Assessments and Evaluations

December 7 & 8, 2006

Page 8Session 101 – Beyond Smile Sheets: Learning Audits and Evaluations We Can Trust, Will Thalheimer, Work-Learning Research

15

© Copyright 2006 Work-Learning Research, Inc.

Learning Versus Retrieval

Search

Cuefosse

ditchRehearsal

Action

Learning

Retrieval

fosse

ditchRehearsal

Action

16

© Copyright 2006 Work-Learning Research, Inc.

Learner Fulfillment

Business Results

Learning to Performance/Results

Learner Retrieves

Learner Applies

Learner Learns

Learning Performance Results

Designing & Developing Online Assessments and Evaluations

December 7 & 8, 2006

Page 9Session 101 – Beyond Smile Sheets: Learning Audits and Evaluations We Can Trust, Will Thalheimer, Work-Learning Research

17

© Copyright 2006 Work-Learning Research, Inc.

Learner Fulfillment

Business Results

Transfer Breakdown

Learner Retrieves

Learner Applies

Learner Learns

Learning Performance Results

18

© Copyright 2006 Work-Learning Research, Inc.

To be a good predictor of on-the-job retrieval, where should a learning evaluation be given?

A. In the training room/context.

B. In a different room/context.

C. It depends...

Designing & Developing Online Assessments and Evaluations

December 7 & 8, 2006

Page 10Session 101 – Beyond Smile Sheets: Learning Audits and Evaluations We Can Trust, Will Thalheimer, Work-Learning Research

19

© Copyright 2006 Work-Learning Research, Inc.

Tested in the learning room or in a different room. Does this matter?

Smith, S. M., Glenberg, A., & Bjork, R. A. (1978). Environmental context and human memory. Memory & Cognition, 6, 342-353.

0

5

10

15

20

25

Tested ina

DifferentRoom

Tested inthe

LearningRoom

20

© Copyright 2006 Work-Learning Research, Inc.

To be a good predictor of on-the-job retrieval, when should a learning evaluation be given?

A. At the end of training.

B. After a delay.

C. At end—and after a delay.

D. It depends...

Designing & Developing Online Assessments and Evaluations

December 7 & 8, 2006

Page 11Session 101 – Beyond Smile Sheets: Learning Audits and Evaluations We Can Trust, Will Thalheimer, Work-Learning Research

21

© Copyright 2006 Work-Learning Research, Inc.

Time Sequence of How Learning Becomes Performance

2 weeks

Retention

On-the-jobPerformance

Performance

1st Event

2nd Event

3rd Event

Learning

22

© Copyright 2006 Work-Learning Research, Inc.

Retrieval Curves

0

20

40

60

80

100

1 2 3 4 5 6 7 8 9 1 0 1 1 1 2 1 3 1 4 1 5

Retention Performance

Posttest

RememberingPretest

Learning

Designing & Developing Online Assessments and Evaluations

December 7 & 8, 2006

Page 12Session 101 – Beyond Smile Sheets: Learning Audits and Evaluations We Can Trust, Will Thalheimer, Work-Learning Research

23

© Copyright 2006 Work-Learning Research, Inc.

Without Workplace Learning

1 2 3 4 5 6 7 8 9 1 0 1 1

LearningForgetting

24

© Copyright 2006 Work-Learning Research, Inc.

With Workplace Learning

1 2 3 4 5 6 7 8 9 1 0 1 1

Learning

WorkplaceLearning

Designing & Developing Online Assessments and Evaluations

December 7 & 8, 2006

Page 13Session 101 – Beyond Smile Sheets: Learning Audits and Evaluations We Can Trust, Will Thalheimer, Work-Learning Research

25

© Copyright 2006 Work-Learning Research, Inc.

With Workplace Learning

1 2 3 4 5 6 7 8 9 1 0 1 1

Learning

WorkplaceLearning

Forgetting

26

© Copyright 2006 Work-Learning Research, Inc.

Which group will do better on the posttests?

Learner LearnsPretest Posttest Delayed

Posttest

Learner Learns

Posttest DelayedPosttest

Group 1

Group 2

Assume:(1) pretest covers same material as posttests;

(2) learners get NO feedback on pretest.

Designing & Developing Online Assessments and Evaluations

December 7 & 8, 2006

Page 14Session 101 – Beyond Smile Sheets: Learning Audits and Evaluations We Can Trust, Will Thalheimer, Work-Learning Research

27

© Copyright 2006 Work-Learning Research, Inc.

Research Example

40

50

60

70

Pressley, M., Tanenbaum, R., McDaniel, M. A., & Wood, E. (1990). What happens when University students try to answer prequestions that accompany textbook material?Contemporary Educational Psychology, 15, 27-35.

No Pre-Questions

Pre-Questions

Read

Pre-QuestionsAnswered

Wrong

Pre-QuestionsAnsweredCorrectly

Pre-Questions Answered

28

© Copyright 2006 Work-Learning Research, Inc.

Tests should be authentic

Level A – Real-world

Level B – High-fidelity simulation

Level C – Scenarios

Level D – Memorization

Level E – Attendance

Level F – Affiliation

Shrock & Coscarelli, Criterion Referenced Test Development

Quasi-Certification

Certification

“Level C represents the last level of certification that can be considered to assess an ability to perform on the job.Level D represents the first quantum jump away from

fidelity in assessment and should be used with caution.”

Designing & Developing Online Assessments and Evaluations

December 7 & 8, 2006

Page 15Session 101 – Beyond Smile Sheets: Learning Audits and Evaluations We Can Trust, Will Thalheimer, Work-Learning Research

29

© Copyright 2006 Work-Learning Research, Inc.

Scenario-based Question

You’re a trainer at a large multinational firm and are

designing a course to teach new-hire high school

graduates word processing skills. You understand the

value of practice and want to create meaningful

practice exercises for your class. For example, for your

beginner class you want them to practice how to set

the margins. How should you design your practice

examples to create the best on-the-job remembering

of the concepts you will teach?

What do you think about this question?

Designing & Developing Online Assessments and Evaluations

December 7 & 8, 2006

Page 16Session 101 – Beyond Smile Sheets: Learning Audits and Evaluations We Can Trust, Will Thalheimer, Work-Learning Research

31

© Copyright 2006 Work-Learning Research, Inc.

What do you think about this question?

32

© Copyright 2006 Work-Learning Research, Inc.

Legal Issues

1. Tests must validly predict job performance, and not create unfair disadvantages.

2. Reliable.

3. You may have to prove your test is valid.

A. Documentation.

B. Post-hoc validation is rarely upheld in court.

C. Starts with job analysis

Designing & Developing Online Assessments and Evaluations

December 7 & 8, 2006

Page 17Session 101 – Beyond Smile Sheets: Learning Audits and Evaluations We Can Trust, Will Thalheimer, Work-Learning Research

33

© Copyright 2006 Work-Learning Research, Inc.

How would you design authentic tests?

1. Test for person who learned MS Excel.

2. Test for person taking leadership class.

3. Test for person learning hair styling.

4. Test for person learning new-product knowledge in order to sell the product.

5. Test for restaurant allergy safety for wait staff.

6. Test for onboarding training.

34

© Copyright 2006 Work-Learning Research, Inc.

Tests can produce learning benefits!

0

1

2

3

4

Nungester, R. J., & Duschastel, P. C. (1982). Testing versus review: Effects on retention. Journal of Educational Psychology, 74, 18-22.

Learn-Learn

Learn-Test

Designing & Developing Online Assessments and Evaluations

December 7 & 8, 2006

Page 18Session 101 – Beyond Smile Sheets: Learning Audits and Evaluations We Can Trust, Will Thalheimer, Work-Learning Research

35

© Copyright 2006 Work-Learning Research, Inc.

Do our assessments validly measure both learning and forgetting?

Completion.

Smile sheets.

Measures of learning tend toward the measurable, not the authentic.

Immediately at end of learning.

In the same room/context as learning took place.

36

© Copyright 2006 Work-Learning Research, Inc.

Most Assessments We Do Cannot Provide Valid Feedback on Our Effectiveness!!

0 20 40 60 80 100

Reaction

Learning

Behavior

Results

ASTD’s industry reports

1.

2.

3.

4.

79.5 %

35.0 %

15.3 %

8.8 %

Designing & Developing Online Assessments and Evaluations

December 7 & 8, 2006

Page 19Session 101 – Beyond Smile Sheets: Learning Audits and Evaluations We Can Trust, Will Thalheimer, Work-Learning Research

37

© Copyright 2006 Work-Learning Research, Inc.

What doesWhat doesthis mean?this mean?

What does this mean?

Instructional Design & Development Process

Feedback for future efforts1 2 3 4

Feedback for current project1 2

Research-Based Instructional Design

Experience-Based Instructional Design Expert-Based Instructional Design

Common-Sense-Based Instructional Design

The Instructional Design & Development Process

Design&

Develop

Deploy&

Evaluate

Prepare,Analyze,

Set Goals

Knowledge,Wisdom,Biases

© Copyright 2006 Work-Learning Research, Inc.

Designing & Developing Online Assessments and Evaluations

December 7 & 8, 2006

Page 20Session 101 – Beyond Smile Sheets: Learning Audits and Evaluations We Can Trust, Will Thalheimer, Work-Learning Research

39

© Copyright 2006 Work-Learning Research, Inc.

Questions

What questions do you have now?

Questions later:

Will ThalheimerWork-Learning Research, Inc.Somerville, Massachusetts, USAinfo@work-learning.com617-718-0067www.work-learning.comwww.LearningAudit.com