Evaluating HRD Programs

Effectiveness

• The degree to which a training (or other HRD program) achieves its intended purpose

• Measures are relative to some starting point

• Measures how well the desired goal is achieved

Evaluation

HRD Evaluation

Textbook definition:

“The systematic collection of descriptive and judgmental information necessary to make effective training decisions related to the selection, adoption, value, and modification of various instructional activities.”

In Other Words…

Are we training:

• the right people

• the right “stuff”

• the right way

• with the right materials

• at the right time?

Evaluation Needs

• Descriptive and judgmental information needed– Objective and subjective data

• Information gathered according to a plan and in a desired format

• Gathered to provide decision making information

Purposes of Evaluation

• Determine whether the program is meeting the intended objectives

• Identify strengths and weaknesses• Determine cost-benefit ratio• Identify who benefited most or least• Determine future participants• Provide information for improving HRD

programs

Purposes of Evaluation – 2

• Reinforce major points to be made

• Gather marketing information

• Determine if training program is appropriate

• Establish management database

Evaluation Bottom Line

• Is HRD a revenue contributor or a revenue user?

• Is HRD credible to line and upper-level managers?

• Are benefits of HRD readily evident to all?

How Often are HRD Evaluations Conducted?

• Not often enough!!!

• Frequently, only end-of-course participant reactions are collected

• Transfer to the workplace is evaluated less frequently

Why HRD Evaluations are Rare

• Reluctance to having HRD programs evaluated

• Evaluation needs expertise and resources

• Factors other than HRD cause performance improvements – e.g.,– Economy– Equipment– Policies, etc.

Need for HRD Evaluation

• Shows the value of HRD

• Provides metrics for HRD efficiency

• Demonstrates value-added approach for HRD

• Demonstrates accountability for HRD activities

• Everyone else has it… why not HRD?

Make or Buy Evaluation

• “I bought it, therefore it is good.”• “Since it’s good, I don’t need to post-test.”• Who says it’s:

– Appropriate?– Effective?– Timely?– Transferable to the workplace?

Evolution of Evaluation Efforts

1. Anecdotal approach – talk to other users

2. Try before buy – borrow and use samples

3. Analytical approach – match research data to training needs

4. Holistic approach – look at overall HRD process, as well as individual training

Models and Frameworks of Evaluation

• Table 7-1 lists six frameworks for evaluation

• The most popular is that of D. Kirkpatrick:

– Reaction– Learning– Job Behavior– Results

Kirkpatrick’s Four Levels

• Reaction– Focus on trainee’s reactions

• Learning– Did they learn what they were supposed to?

• Job Behavior– Was it used on job?

• Results– Did it improve the organization’s effectiveness?

Issues Concerning Kirkpatrick’s Framework

• Most organizations don’t evaluate at all four levels

• Focuses only on post-training

• Doesn’t treat inter-stage improvements

• WHAT ARE YOUR THOUGHTS?

A Suggested Framework – 1

• Reaction– Did trainees like the training?– Did the training seem useful?

• Learning– How much did they learn?

• Behavior– What behavior change occurred?

Suggested Framework – 2

• Results– What were the tangible outcomes?– What was the return on investment

(ROI)?– What was the contribution to the

organization?

Data Collection for HRD Evaluation

Possible methods:

• Interviews

• Questionnaires

• Direct observation

• Written tests

• Simulation/Performance tests

• Archival performance information

Interviews

Advantages:• Flexible• Opportunity for

clarification• Depth possible• Personal contact

Limitations:• High reactive effects• High cost• Face-to-face threat

potential• Labor intensive• Trained observers

needed

Questionnaires

Advantages:• Low cost to

administer• Honesty increased• Anonymity possible• Respondent sets the

pace• Variety of options

Limitations:• Possible inaccurate

data• Response conditions

not controlled• Respondents set

varying paces• Uncontrolled return

rate

Direct Observation

Advantages:• Nonthreatening• Excellent way to

measure behavior change

Limitations:• Possibly disruptive• Reactive effects are

possible• May be unreliable• Need trained

observers

Written Tests

Advantages:• Low purchase cost• Readily scored• Quickly processed• Easily administered• Wide sampling

possible

Limitations:• May be threatening• Possibly no relation to

job performance• Measures only

cognitive learning• Relies on norms• Concern for racial/

ethnic bias

Simulation/Performance Tests

Advantages:• Reliable• Objective• Close relation to job

performance• Includes cognitive,

psychomotor and affective domains

Limitations:• Time consuming• Simulations often

difficult to create• High costs to

development and use

Archival Performance Data

Advantages:• Reliable• Objective• Job-based• Easy to review• Minimal reactive

effects

Limitations:• Criteria for keeping/

discarding records• Information system

discrepancies• Indirect• Not always usable• Records prepared for

other purposes

Choosing Data Collection Methods

• Reliability– Consistency of results, and freedom from

collection method bias and error

• Validity– Does the device measure what we want to

measure?

• Practicality– Does it make sense in terms of the resources

used to get the data?

Type of Data Used/Needed

• Individual performance

• Systemwide performance

• Economic

Individual Performance Data

• Individual knowledge

• Individual behaviors• Examples:

– Test scores– Performance quantity, quality, and timeliness– Attendance records– Attitudes

Systemwide Performance Data

• Productivity

• Scrap/rework rates

• Customer satisfaction levels

• On-time performance levels

• Quality rates and improvement rates

Economic Data

• Profits• Product liability claims• Avoidance of penalties• Market share• Competitive position• Return on investment (ROI)• Financial utility calculations

Use of Self-Report Data

• Most common method• Pre-training and post-training data • Problems:

– Mono-method bias• Desire to be consistent between tests

– Socially desirable responses

– Response Shift Bias: • Trainees adjust expectations to training

Research Design

Specifies in advance:

• the expected results of the study

• the methods of data collection to be used

• how the data will be analyzed

Research Design Issues

• Pretest and Posttest– Shows trainee what training has

accomplished– Helps eliminate pretest knowledge bias

• Control Group– Compares performance of group with training

against the performance of a similar group without training

Recommended Research Design

• Pretest and posttest with control group

• Whenever possible:– Randomly assign individuals to the test group

and the control group to minimize bias– Use “time-series” approach to data collection

to verify performance improvement is due to training

Ethical Issues Concerning Evaluation Research

• Confidentiality

• Informed consent

• Withholding training from control groups

• Use of deception

• Pressure to produce positive results

Assessing the Impact of HRD

• Money is the language of business.• You MUST talk dollars, not HRD jargon.• No one (except maybe you) cares about

“the effectiveness of training interventions as measured by and analysis of formal pretest, posttest control group data.”

HRD Program Assessment

• HRD programs and training are investments

• Line managers often see HR and HRD as costs – i.e., revenue users, not revenue producers

• You must prove your worth to the organization – Or you’ll have to find another organization…

Evaluation of Training Costs

• Cost-benefit analysis– Compares cost of training to benefits gained

such as attitudes, reduction in accidents, reduction in employee sick-days, etc.

• Cost-effectiveness analysis– Focuses on increases in quality, reduction in

scrap/rework, productivity, etc.

Return on Investment

• Return on investment = Results/Costs

Calculating Training Return On Investment

    Results Results    

Operational How Before After Differences ExpressedResults Area Measured Training Training (+ or –) in $

Quality of panels % rejected 2% rejected 1.5% rejected

.5% $720 per day

    1,440 panels 1,080 panels

360 panels

$172,800      per day   per day     per year

Housekeeping Visual 10 defects 2 defects 8 defects

Not measur-    inspection   (average)   (average)     able in $

   using

  

   

 

  20-item

       

 

  checklist

       

Preventable Number of 24 per year 16 per year 8 per year

   accidents   accidents

       

  Direct cost $144,000 $96,000 per

$48,000 $48,000 per

   of each   per year   year

   year

   accident

       

 

 

Return

Investment

    Total savings: $220,800.00

ROI = =    

   

   

   SOURCE: From D. G. Robinson & J. Robinson (1989). Training for impact. Training and Development Journal, 43(8), 41. Printed by

permission.

Operational Results Training Costs

= $220,800$32,564

= 6.8

Types of Training Costs

• Direct costs

• Indirect costs

• Development costs

• Overhead costs

• Compensation for participants

Direct Costs

• Instructor– Base pay– Fringe benefits– Travel and per diem

• Materials• Classroom and audiovisual equipment• Travel • Food and refreshments

Indirect Costs

• Training management

• Clerical/Administrative

• Postal/shipping, telephone, computers, etc.

• Pre- and post-learning materials

• Other overhead costs

Development Costs

• Fee to purchase program

• Costs to tailor program to organization

• Instructor training costs

Overhead Costs

• General organization support

• Top management participation

• Utilities, facilities

• General and administrative costs, such as HRM

Compensation for Participants

• Participants’ salary and benefits for time away from job

• Travel, lodging, and per-diem costs

Measuring Benefits

– Change in quality per unit measured in dollars– Reduction in scrap/rework measured in dollar

cost of labor and materials– Reduction in preventable accidents measured

in dollars– ROI = Benefits/Training costs

Utility Analysis

• Uses a statistical approach to support claims of training effectiveness:– N = Number of trainees– T = Length of time benefits are expected to last– dt = True performance difference resulting from

training– SDy = Dollar value of untrained job performance (in

standard deviation units)– C = Cost of training

U = (N)(T)(dt)(Sdy) – C

Critical Information for Utility Analysis

• dt = difference in units between trained/untrained, divided by standard deviation in units produced by trained

• SDy = standard deviation in dollars, or overall productivity of organization

Ways to Improve HRD Assessment

• Walk the walk, talk the talk: MONEY

• Involve HRD in strategic planning

• Involve management in HRD planning and estimation efforts– Gain mutual ownership

• Use credible and conservative estimates

• Share credit for successes and blame for failures

HRD Evaluation Steps

1. Analyze needs.

2. Determine explicit evaluation strategy.

3. Insist on specific and measurable training objectives.

4. Obtain participant reactions.

5. Develop criterion measures/instruments to measure results.

6. Plan and execute evaluation strategy.

Summary

• Training results must be measured against costs

• Training must contribute to the “bottom line”

• HRD must justify itself repeatedly as a revenue enhancer, not a revenue waster