Performance SF 2001

HUMAN PERFORMANCE TECHNOLOGY: RESEARCH AND THEORY TO PRACTICE

Harold D. Stolovitch

Human performance technology (HPT) has made impressive strides since it first became a term and an emerging field of practice in the 1970s. The dawn of the 21st century finds HPT flourishing. The literature has expanded fivefold in the last 10 years; the number of professional practitioners has multiplied at an even greater rate; the list of academic institutions offering graduate courses as well as full degree programs has become impressive. However, along with accelerated growth and presence, a deep concern has emerged that the practice of HPT is outstripping its theory and research base.

This article focuses on ensuring that HPT practices are clearly founded on sound theoretical constructs, scientifically derived evidence and respectable, well-documented professional precedent.

HPT practitioners generally base their interventions on principles they consider fundamental truths. These drive their decision-making and professional activities. What follows is an examination of eight common myths with commentary based on research and theory. Discussion of each concludes with keys for practice drawn from the research.

Feedback leads to improved performance.
This is one of the most frequently cited principles of HPT. It seems intuitively sensible. HPT practitioners often cite anecdotes to support this statement. Most attribute the scientific origins of this principle to Thorndike's Law of Effect (1913) which says that when a correct response is reinforced by a reward (for example, "That's correct!"), the reward strengthens the bond between the environment and the appropriate response - somehow equating feedback with reinforcement. However, Thorndike noted that feedback interventions can have detrimental properties if viewed in comparison with others or if vague. A large body of research on knowledge of results, knowledge of performance and feedback interventions implies that feedback given to a person who is learning or carrying out a task results in performance improvement.

So how does one explain the following two citations?

"Feedback does not uniformly improve performance" (Balcazar et al., 1985).
"Few concepts in psychology have been written about more uncritically and incorrectly than that of feedback…feedback is information, that is, data, and as such has no necessary consequences at all" (Latham & Locke, 1991).

These affirmations from serious researchers raise doubts about the generalization that feedback leads to improved performance.

Raising even greater uncertainties is the comprehensive research review on the effects of feedback by Kluger and DeNisi (1996) who rigorously examined 2,500 studies dating back to the 1890s on feedback and its effects on learning and performance. They included 607 effect sizes and 23,663 observations. They conclude that the question - "Does feedback improve performance?" - is an inappropriate one. More suitable is "What must one do so that feedback improves performance?" Their key point is that there is a need for a consistent and comprehensive theory of feedback to support actions. They suggest that this theory must account for three classes of variables:

Feedback cues. These can shift the learner's/performer's attention to or from a task. They can debilitate performance if viewed as being directed at self, as "controlling," or as coming from an external source that does not seem to value the individual. Feedback cues perceived as motivational (emphasizing gains, accepted as corrective and helpful or emphasizing benefits) help improve performance. Specific feedback can improve performance by targeting clearly identified behaviors. It also increases cognitive load through added detail.
Task. The fewer the cognitive resources required for task performance, the more positive the effect of the feedback on performance (i.e. the simpler the task for the performer, the greater the impact of the feedback).
Situational. Goal setting increases the feedback effect on performance. The clearer and more valued the goal, the greater the impact on performance. Personality variables affect impact. Lower self- esteem, greater externality of locus of control or greater susceptibility to cognitive interference lessens the effect of feedback.

Keys for practice from the research evidence are: feedback on process and successful outcomes improves performance; has more effect on cognitive than physical tasks; has negative effects if it threatens self-esteem; has a stronger effect on memory than rule-following tasks.

What appears evident is that blanket statements such as "feedback leads to improved performance" lack nuance and can lead to inappropriate decision-making.

Immediate feedback is more effective in improving performance than delayed feedback.
While this also seems to make good sense, numerous studies contradict it.

In one study, trainers taught communicative gestures to individuals with severe retardation. Duke (1995) discovered that delayed feedback resulted in increased training accuracy. Studying cognitive growth and development, van Geert (1991) demonstrated that these depend more on delayed feedback. Kern-Dunlap (1992) found that delayed feedback and reinforcement improved desirable peer interactions. Sasaki (1997) uncovered numerous instances in which delayed feedback produced superior performance results. Focusing on working memory constraints, Sasaki suggests that immediate feedback can create overload producing an inhibitory effect that lessens feedback impact.

The research and theory literature offers two keys for practice. For simple neutral tasks, use immediate feedback. For complex tasks, apply delayed feedback to obtain superior results.

Job satisfaction generally leads to improved performance.
Happy workers work harder and do better. While intuitively sensible, this
may lead to erroneous decisions.

While training may often be necessary, rarely is it sufficient to achieve sustained improved post-training performance.

Macan (1996) studied time management training, effects on work attitudes and job performance. He concluded that there were no main effects or interactions found for job satisfaction. Fried, in an extensive meta-analytic study on work satisfaction and performance, found "dissimilarity…in relationships of…
performance and…satisfaction…particularly substantial" (1991, p.690). More bluntly, Mannheim et al. (1997) concluded in their study of personnel in the high-technology industry that job satisfaction and performance are not significantly correlated. Iaffaldano and Muchinsky (1985) determined that the satisfaction-performance relation constitutes an illusory correlation and discovered that information on job satisfaction can significantly bias performance appraisals.

Lee (1992) neatly summarizes the job satisfaction-performance literature with four useful keys for HPT practice. To obtain improved performance, provide effective management of work patterns; encourage workers; support and reward more intensive work efforts; provide job challenge.

Successful performance during training usually results in improved post-training performance.
If individuals perform well during training, shouldn't it carry over to the job? If not, why train? While training may often be necessary, rarely is it sufficient to achieve sustained improved post-training performance. Convincing organizations to assume a broader performance perspective is a constant challenge. This, despite findings such as those of Baldwin and Ford (1988), who, in an extensive review of training research, found that although "…American industries annually spend more than $100 billion on training…not more than 10% of these expenditures actually result in transfer to the job." Ford and Weissbein (1997) found that nine years later, transfer results remained unchanged. Esque and McCausland (1997) reported that at Intel, a popular, highly rated management course taken by 600 managers, resulted in less than one percent transfer. Without adequate management attention, performance deteriorates rapidly following training - often to below pretraining performance levels - due to non-supportive or even inhibitory factors.

Drawing from research and theory literature, Stolovitch and Maurice (1998) found clear guidelines and considerations for practice to improve the impact of training on performance:

Ensure that those chosen to participate in training are appropriate for the training.
Prior to training, provide briefings to participants on expected post-training application.
Design and deliver training according to learner characteristics and performance needs.
Provide immediate supervision and support, encouragement, and resources post-training.
Provide consistent, specific, corrective and confirming feedback to improve and maintain performance.
Offer personally meaningful benefits to incite learners to persist with the new skills and knowledge.

If you want to learn how to do something, go to the expert.
The field of knowledge engineering has discovered that experts cannot articulate the knowledge they use when demonstrating their expertise. This conforms well to a substantial body of theory and research. As Winkles (1992) points out, experts can tell what they do in specific instances but cannot recommend general principles that apply in all cases.

Cognitive psychological research and theory suggest key differences between procedural knowledge (able to do, know how) and declarative knowledge (able to talk about, know what). While experts perform well, they generally are inarticulate about what they are doing (unconscious competence) and experience great difficulty explaining their performance to novices.

Experts are necessary to provide input into training and to build credibility and depth in content. However, research suggests focusing on other factors to create effective learning.

Two key principles for HPT practice from research and theory are:

Do not go to the expert to learn how to do something.
Draw from the expert as one important input source for engineering the learning system.

Physical capital generates a significantly higher return-on-investment (ROI) than human capital.
Conventional wisdom suggests that substantial, strategic investments in plant, equipment and technology ensure market competitiveness and higher ROI yields. Yet studies such as those conducted by Huselid (1995) and Huselid and Becker (1997) clearly demonstrate the higher value of investment in human capital. Pfeffer (1998) cites numerous studies from virtually every industrial sector and continent that demonstrate astonishingly high ROI rates from "high-performance management practices." Nobel prize-winning economists Shultz (1981) and Becker (1993) demonstrated incredibly large returns that human capital investment generates at macroeconomic levels. Stewart (1997) and Phillips (1997) have documented the ROI in human and intellectual capital on microeconomic scales. Edvinsson and Malone (1997) estimate that the median book value of US corporations undervalues market worth by about 40 percent. For knowledge-based companies, the median missing value on the balance sheet is in excess of 100 percent. Bradley (1996) calculated acquisitions to book value of US companies to be about 4:1 and for knowledge-based companies, 10+:1. When IBM purchased Lotus, the acquisition price of $3.5 billion exceeded the book value of $230 million by a ratio of 15:1 (Oliver, 1995).

As key for practice, the HPT practitioner must demonstrate the bottom-line value of investing in human performance improvement initiatives. The research conclusively demonstrates that well-designed human performance systems leverage existing human capital to generate high value compared to cost.

Technology advances since 1970 have consistently accelerated an overall increase in work productivity.
Until 1998, nothing was further from the truth. Recently, investments in technology have shown significant productivity gains, particularly in the manufacturing sector. However, despite the increased industry investment in information technology of more than 600 percent between 1978 and 1996, no concurrent gains in productivity were recorded (Slater & Strawser, 1997). Currently, the increases noted from 1998-2000 have again begun to decline despite high technology investment (Galvin, 2001).

Research and statistical findings comparing enthusiasm for technology with productivity gains raise cautions for HPT practitioners. Key for practice is that HPT practitioners keep themselves constantly informed of technological advances, ready to accept demonstrable gains in efficiency and effectiveness, but cautious in counseling clients about anticipated performance improvements. Balance technology enthusiasm with respected, credible data that indicate a high probability for replication of improved performance results.

Common sense is a friendly ally of science.
Most research methodology books warn research students that "common sense" presents one of the greatest dangers to scientific truth and is frequently an enemy of science. Common sense has often been a major cause of suppression of scientific truth (i.e., the belief in a heliocentric universe; maintenance of practices based on a flat Earth, gender oppression and racial injustice).

Common sense provides subjective meaning to facts (often sparse, poorly defined and collected in a biased manner). Common sense tends to overgeneralize (from a few isolated incidents to an entire population). It also is driven by logical and illogical extrapolations. As White discovered in his investigations and experimentation on common-sense constructions of causal processes: "Studies on how humans create cause-effect relationships demonstrate that they tend to draw linear rather than systemic conclusions and are resistant to complex, data-based explanations of phenomena" (1955, p. 377).

For HPT practitioners, the key messages for practice that emerge from the research are clear:

Do not trust common sense.
Be data driven.
Be theory driven - theory being the scientific explanation of phenomena backed by solid, empirical research evidence.

The purpose of this article has been to encourage HPT practitioners to base their discussion and practice on sound theoretical constructs, scientifically derived evidence and respectable, well-documented professional precedent.

Theory and research add depth and weight to complex issues and nourish successful HPT practice. If you are a practitioner, think of yourself every day as a researcher. If you are a researcher, think of yourself every day as a practitioner. Through respect and adherence to our scientific roots, we can ensure that HPT will maintain its right to endure.

All references for this article are located in the complete article on HSA’s Website at: www.hsa-lps.com/Articles.htm

This article is an excerpt from the original award-winning article that was published in the International Society for Performance Improvement’s Performance Improvement Journal, Vol. 39, No. 4.