Classical Conditioning Measures and Contingency

In classical conditioning studies, a neutral stimulus and a biologically significant stimulus (US) are presented close in time.
The US elicits a response, which makes it difficult to determine the response associated with the neutral stimulus.
Determining when the response begins can help ascertain if the neutral stimulus has a response associated with it.
- If the response begins during the neutral stimulus, before the US presentation, it likely indicates conditioning.

Learning can be measured in terms of latency, which is the time at which a response occurs.
However, if the interval between the conditioned stimulus (CS) and the US is short, measuring the response to the CS becomes difficult.

To measure conditioning when intervals are short, test trials are used.
Test trials involve presenting the CS alone, without the US.
Test trials are less frequent than conditioning trials to avoid reducing conditioning.
Learning is measured by the frequency of the conditioned response during these test trials.
- Example: Eye blink conditioning where a tone (CS) is paired with an air puff (US), and the blink response to the tone is measured.

Intensity or size of the response: A larger startle response indicates greater fear.
Duration of the response: How long a rat freezes (stays still) indicates fear.

Repeated presentation of an intense, biologically significant stimulus can lead to sensitization.
Example: Repeated electric shocks can make a person more aroused and responsive to everything.
Increased responding may appear to be from pairing a neutral stimulus with the shock, but it is due to repeated shocks.

Sensitization can lead to pseudo conditioning, where an initially neutral stimulus elicits a response because it is presented in a context where the response has been sensitized.
Example: A startle response occurs due to fear of shocks, making anything presented in the shock context elicit a startle.

Conditioning occurs only with explicit pairing between the CS and the US.
Random presentation of the CS and the US will not result in conditioning but can cause sensitization and pseudo conditioning.

Comparing responding in a conditioning group (CS and US paired) with a control group (random CS and US presentations) helps determine the extent of conditioning.
If the conditioning group shows a higher level of responding than the control group, the increased response can be attributed to conditioning.

Differential conditioning involves pairing one cue with the US and presenting another neutral cue without the US in the same context.
Evidence for conditioning is shown by a greater response to the stimulus paired with the US compared to the stimulus never paired with the US.
Example: One picture is always followed by a shock, while another picture is never followed by a shock. Responses to the two stimuli are compared.

In the real world, the relationship between stimuli is not always regular.
Classical conditioning depends on the degree of contingency between the CS and the US.
Contingency is the extent to which the CS and the US occur together and do not occur on their own.

The strength of contingency is defined as: $P(US|CS) - P(US|no CS)$
- Where $P(US|CS)$ is the probability of the US occurring given the presence of the CS
- and $P(US|no CS)$ is the probability of the US occurring given the absence of the CS.
Example: If the CS is the house next door and the US is encountering a scary dog, the contingency is the probability of seeing the scary dog at the house next door minus the probability of seeing the scary dog somewhere else.

In the standard lab setting, the US always occurs during the CS and never without it.
$P(US|CS) = 1$ and $P(US|no CS) = 0$
This results in strong conditioning.

Sometimes the CS is reinforced by the US, other times it is not.
Example: Sometimes there is a scary dog at the house next door; other times, the dog is asleep.
If the scary dog is present half the time, $P(US|CS) = 0.5$ . If scary dogs are never seen elsewhere, $P(US|no CS) = 0$
This also results in relatively robust conditioning.

A random CS-US relationship means the US occurs randomly, independent of the CS.
$P(US|CS)$ is the same as $P(US|no CS)$ , so the contingency is zero.
Conditioning does not generally occur under these conditions.
Example: Seeing the scary dog half the time at the house next door and half the time elsewhere means $P(US|CS) = 0.5$ and $P(US|no CS) = 0.5$ . This leads to a contingency of zero.

With a negative contingency, the neutral stimulus predicts the absence of the biologically significant stimulus, leading to inhibitory learning.
Inhibitory learning is when the CS prevents the occurrence or reduces the size of a conditioned response.
The US must be expected to occur for its absence to be predictable by the CS.

A cranky old man yells at you when you ride your bike or come to your door.
When your mom is present, the old man leaves you alone, creating a negative contingency.
Fear is present all the time except when your mom is with you.
Your mom acts as an inhibitory stimulus, inhibiting your fear response.

When you deliver a paper to the house down the street, the cranky old man yells at you. However, when your mom is with you and you deliver meals on wheels to the house, his wife answers the door and she's sweet as pie.
You learn to be afraid when going to the house alone but not with your mom.
Your mom acts as an inhibitory stimulus for your fear.

Present the inhibitory stimulus with a separately trained CS that produces a conditioned response.
If you are afraid of the house next door because of a scary dog, you should feel less afraid if your mom, who has been trained as an inhibitor of fear, is with you.
This is based on the idea that conditioned inhibition counteracts a conditioned response.

If a stimulus actively inhibits a particular response, it should be difficult to condition that stimulus to elicit the response.
If your doctor gives you a painful injection every time you see him, you're likely to quickly learn to be afraid of going to the doctor. However, if your mother who makes you feel safe starts giving you painful injections, it's likely to take longer to learn to be afraid of her
Inhibitory conditioning takes longer because the feeling of safety must be overcome.