51. Variable Rewards are Powerful

● Fixed interval:

Reinforcement based on time intervals is always the same.

● Variable interval:

Reinforcement is based on time. The amount of time varies, but it

averages to a particular time.

● Fixed ratio:

Reinforcement is based on the number of bar presses and the number is

always the same.

● Variable ratio:

Reinforcement is based on the number of bar presses. The number

varies, but it averages to a particular ratio.

If you're not sure that operant conditioning is related to design, think about it more deeply. Many

times as designers you want to encourage a certain behavior continuously.

Operant Conditioning and Design

A guy named BF Skinner conducted an experiment on a rat. The study was to study the rats behavior after experiencing positive reinforcement.

Terms

Skinner tested out various scenarios, and found that

the frequency with which you gave the food pellet, and whether you give it based on elapsed-time or bar presses, affected how often the rat would press the bar.

Interval schedules. You provide a food pellet after a certain interval of time has passed,

for example, 5 minutes. The rat gets a food pellet the first time he presses the bar after 5

minutes is up.

Ratio schedules. Instead of basing the reinforcements on time, you base it on the

number of bar presses. The rat gets a food pellet after every 10 bar presses

There's another twist. You can have fixed or variable variations on each schedule. If it's a fixed

schedule, then you keep the same interval or ratio, for example, every 5 minutes for every 10

presses. If it's variable, then you vary the time or ratio, but it averages out; for example,

sometimes you provide the reinforcement after 2 minutes, sometimes after 8 minutes, but it

averages out to 5 minutes. So altogether there are four possible schedules:

TAKEAWAYS

For operant conditioning to work, the reinforcement (reward) must be something that particular

audience wants. Hungry rats want food pellets. What is your particular audience really want?

Think about the pattern of behavior you're looking for and then adjust the schedule of rewards to

fit that schedule. Use a variable ratio schedule for the maximum Behavior repetition.

Citation Reference: “Chapter 6 / Section 51.” 100 Things Every Designer Needs to Know about People, by Susan Weinschenk, New Riders, 2011, pp. 118–120.