This report begins with a background to behaviour modification, including details on the psychological mechanisms and the reasons behind the theory. These ideas are then taken further and applied within an organisational setting using published examples, followed by a general appraisal on the work done.
In the late 1890s, Ivan Petrovich Pavlov experimented on a dog to view its reaction to having a ringing bell associated with getting fed (this type of association is called a trial). Pavlov discovered that after “training”, the dog would involuntarily salivate to the sound of the bell, regardless of the food’s presence (see figure 1). The experimenters first performed a minor operation on a dog to relocate its salivary duct to the outside of its cheek, so that drops of saliva could be more easily measured. The dog, which was food deprived, was then harnessed in an apparatus to keep it steady. Periodically, a tone was sounded, followed shortly thereafter by meat powder being placed in the hungry dog’s mouth. Initially, the dog shows little responsiveness to the tone. Over time, however, the dog comes to salivate at the sounding of the tone alone. When this occurs, Pavlovian conditioning has occurred, in that a new, or conditional, reflex has developed.In effect, the dog became conditioned. This is more commonly known as classical conditioning.
Figure 1 – The experimental set up for Pavlov’s conditioning
Classical conditioning works on instinct and reflexes – stimulus (S) leading to response (R), or S->R, and as such has limited applications and use; it is “stimulus bound??? and non-directed.
In 1935, Skinner took classical conditioning one-step further, albeit a very significant one. Skinner identified between two types of behaviour, namely the respondent and the operant. The respondent was basically the classical conditioning mechanism, whereas the operant was an extension on classical conditioning, where behaviour is actually learnt.
Pavlovian conditioning is largely responsible for our motivation to respond in any situation. Operant conditioning, on the other hand, is what we learn to do to satisfy these motivational states. Another way of looking at this difference is to look at the relationship between behaviour and the environment. In Pavlovian conditioning, the relationship is stimulus-response (S->R), but in operant it is response-stimulus (R->S). As seen in figure 2, classical conditioning sets up operant conditioning.
Figure 2 – Stimulus response diagram for operant conditioning. Note how classical conditioning is a prerequisite for operant conditioning
Whether responses occur in the future depend upon the nature of the consequence or more formally, contingency. If it makes life better for the organism, it will likely occur again in the future (this is termed reinforcement), and if it makes like worse, it will likely not occur again in the future (this is called punishment).
These facts can be summed up by Kazdin’s square, shown in Table 1. The table summarises, in a very simplistic way, the basic relationships that our behaviour can have with our environment. In its simplest elements, there are two types of stimuli in the world; those that we like (nice) and those we don’t (nasty). Actions that have an effect can either lead to one of these events either being added to our environment (get) or withdrawn (taken-away). There is one more item that has no place within the square, but also has an effect, namely non-reward.
Positive reinforcement is straightforward, where the individual’s behaviour adds something desirable to the environment, thereby increasing the probability, under similar circumstances, of that behaviour occurring again in the future.
Negative reinforcement is less straightforward, with two types occurring:
- Escape is when something unpleasant is currently going on, and the organism does something to terminate or lessen it, for example when a parent picks up a crying baby – the baby stops crying and the parent’s behaviour of picking the baby is negatively reinforced;
- Avoidance is when behaviour has prevented the onset of something unpleasant occurring. As an example, someone with hey-fever will take anti-histamine a couple of hours before going outside when there is a high pollen count.
Punishment also has two subsets, where something nice is taken away or something nasty is given. An example of the former is when you get fined money for breaking the law (e.g. not wearing a seatbelt). Getting a shock by touching an electric fence is an (extreme!) example of the latter. In both cases, the probability of the same behaviour occurring will decrease, as the outcomes are highly undesirable.
Maintaining behaviour – The effectiveness of each consequence is mainly related to the frequency it follows, or does not follow, each occurrence of the behaviour. This is an issue that leads to the maintenance of a learnt behaviour, and is dealt with variable ratio reinforcement; a subject is more likely to persevere a beneficial behaviour if the reward payoff occurs randomly. This is the mechanism behind gambling and is much more powerful and simple to implement than constant reward.
Suppressing behaviour – There are two options for this, namely punishment and non-reward. For punishment to be effective, it must occur immediately after the behaviour, as well as consistently, which is very difficult. Furthermore, punishment has been shown to merely suppress behaviour and not extinguish it. Non-reward, on the other hand, has been found to be far more effective. By the outcome of the behaviour removed, there is no need for the behaviour to occur. After an initial extinction burst, where the behaviour is amplified, the behaviour dies down and becomes extinct.
Organisational Applications of Behaviour Modification
The above section concentrated on the general theory behind behaviour modification. However, the theory is also very relevant within an organisational setting, and is commonly used to promote a desired behaviour (i.e. number of sales, customer courtesy, safety in the workplace etc.), or to address issues such as motivation and turnover, as the following two examples illustrate in some detail.
Behaviour Management in the Public Accommodations Industry
This example of a scientifically studied and implemented illustration of organisational behaviour modification, or OBMod for short, was carried out with existing hotel cleaning staff. The aim was to thoroughly deep clean and then maintain property cleanliness with minimum turnover and cost.
The reason behind the implementation of this technique was due to the very high turnover rate of cleaning staff, and the substandard work carried out. To further exacerbate management’s situation, the hiring, training, outfitting and maintenance of housekeeping staff is one of the largest budget line item for most hotel/motel operations. Needless to say, most hotels have tried various methods to change the situation, but at the hotel where the OBMod was done, typical of most hotels, had very little success in improving the situation.
In this particular hotel, the following issues were raised in context of the OBMod process:
- A definition of the behaviour of cleanliness would have to be created, that would be relevant to the hotel industry;
- The definition would require a rigorous reliability procedure that would remain compatible with the daily activities of the cleaning staff;
- There was a distinct lack of feedback to the cleaners, hence a feedback system would be required as part of the intervention;
- Aversive managerial practices were common, so a performance-contingent type implementation would help, and coincide with the feedback, increasing low morale.
The last point is a clear use of reinforcement to promote the required behaviour. With the behaviour and requirements defined, the intervention was put into action, and comprised of the following:
- The initial training of the staff measuring the performance, as well as measuring the pre-intervention cleanliness for a period (baseline measurement);
- The prominent and public placement of a chart displaying individual work output per day;
- The original checklist became more specific, pinning down specific tasks and contained 70 points; the cleaners mutually agreed on its contents.
The baseline was measured to be a mere 20% of the defined criteria for “cleanliness???. As a result, it was deemed that the full introduction of the 70-point checklist would pose a shock for the workers, and was thus implemented in staggered stages over time. The cleaners were informed specifically that:
- The checking of the rooms would be random;
- A preset goal was given (i.e., for the first staggered stage to score 18 in the 21-point checklist;
- Participation in the scheme was encouraged, but not required; non-participants would not be penalised;
- Management would train, equip and supply participants, in order for them to attain the required 18 points;
- Target meeting employees would be able to exchange points for time off, dinner out etc.;
- After six months, if all cleaners participated, a vote could be taken to continue or discontinue the scheme.
Over the staggered increments of the 70-point checklist, periods of withdrawal were conducted, where no inference occurred, to measure the effects of the scheme. The data produced displayed the following facts:
- The baseline performance rose to over 90% very quickly;
- The fluctuations in cleanliness were smoothed over in time;
- Despite withdrawal of interference, the new baseline performance remained the same.
The program was considered a success, with management saving money, as well as gaining a marked increase in cleaning performance. However, being rushed as well as heavy perspiration of the cleaners was observed – it was thus concluded that if it were not for heavy incentives, they would not have worked as hard, and the improvement would have been marginal.
Getting Real Estate Agents to Sell More
Whilst improving the sales techniques of agents is of major importance to the companies, training on sales techniques have proved to have no lasting effects, hence an OBMod initiative was carried out in one company. The aim was to improve and maintain in the long-term sales efficiency through improved seller-client relations.
It is enlightening to illustrate the perceived faults of the traditional sales programs, namely:
- Not clearly specifying the required agent behaviour in order to achieve desired company outcomes;
- The lack of measurement systems for agents to monitor progress on a daily basis;
- No agent feedback on performance encouraged;
- The lack of procedure for maintaining and strengthening sales-relevant actions.
In order to measure the baseline, the agents were given a form to fill in over 20 weeks (usually covert measures are undertaken, but this project was based on the measurement of two behaviours that were dependant on agent self-report). The form consisted of general client information, as well as nature of contact and actions taken.
The main correlation that was determined to increase sales was defined as having a large client contract base, hence the following behaviours were outlined to encourage this correlation:
- Increase the initial contact with prospective clients;
- Follow up contacts with established clients;
- Contact was defined as a face-to-face type meeting.
The form allowed the recording of such facts. Since the program depended on self-assessment, a questionnaire was sent to clients who were reported by agents as being met, to discern whether it actually happened.
The results showed that there was a marked improvement on the baseline once the intervention took place, rewarding employees (and hence reinforcing their behaviour) through tokens that were exchangeable for goods, such as pool tables. However, there was a drop towards the end of the intervention period, due to the anticipation of the pre-announced withdrawal. When withdrawal actually took place (the self-monitoring remaining, but the reward system removed), morale and productivity stooped even below the baseline level.
The final stage, coined “reversal??? was the done after discussions with the agents. They all wanted to retain the reward scheme, but decided to put emphasis on the following up of customers, instead of contacting of new customers. When put in place, the productivity increased to values akin to the intervention stage.
The understanding of OBMod is relatively old, and very powerful as the above examples illustrate, yet it is seemingly under-utilised, or more often, done incorrectly. What is very clear is that it should be done scientifically and quantitatively, and not in an ad hoc manner. An example of bad attempts at promoting behaviour is the initial training of the estate agents (outlined above), where little emphasis was placed on defining and maintaining the desired behaviours.
What is crucial is that the behaviour is correctly defined and presented in a measurable manner. Furthermore, organisations should work on the idea that employees want to please, a fact, when fully accepted, will remove aversive management and, for example, promote feedback and variably reinforced praise.
A negative issue raised in both examples was that the program was not specifically geared towards dealing with underachievers, which meant that they were commonly left out. Had the programs provided more encouragement, there would have been a better reaction, as they too don’t want to be left out (social conformity being the main driver in this situation). Furthermore, management should not use OBMod to counter more fundamental issues, such as understaffing – if the cleaners in the first example were constantly out of breath, yet meeting their required 90% cleanliness quota, isn’t that a sign of understaffing?
The final issue is that the subjects need to be aware of the situation, and also want to change; a more covert process would not be as effective, and moral issues are raised.