WSLS and pronouns or similar problems

[Guest guest]

Ray,

thank you for this technical information. It reminds me we may face something

similar with our son Nicolas, 5yo, verbal, when doing pronouns and before/after

programs. Those programs brings Sd with only two possible responses (touch your

nose Vs touch my nose ; before vs after), hence we may easily fall in WSLS and I

think we did.

The result is that we are, for example for the my/your pronouns program, stuck

at the same place for months.

I understand we must change our approach but have not been able to find a

suitable one up to now. Please do you have any idea, you or anybody here on this

list ?

Thank you

Benoit

Message du 15/07/03 00:32

De : cepdti2@...

A :

Copie à :

Objet : [ ] Re: WIN_STAY_LOSE_SHIFT

Hey everyone,

I received an email from Jennica asking me to explain what

win-stay-lose-shift is. Here is a copy of the reply I sent to her:

Hey Jennica,

I really think that is great that you are so active helping other parents,

posting messages etc. please do not ever hesitate to contact me, it may take a

couple of days to return your message, but I will always get back to you. And

go ahead and forward this to or any other appropriate place.

I'll do my best to give an example 'win-stay-lose-shift', here goes:

First, this is most likely to occur in smaller field sizes, say two or three

items when instructors employ the typical discrimination learning procedure

which looks like this:

If SD1 = " cup " and SD2 = " shoe "

1. Present SD1 in isolation with neutral distracters

(NOTE#1: please remember that isolation does not necessarily mean that this

is the ONLY item on the table, by using neutral distracters and only asking for

the target item, the response may still be considered to be in isolation)

2. Present SD2 in isolation

3. Preset SD2 in the presence of SD1 but asking for SD2 only.

4. Present SD1 in the presence of SD2 asking for SD1 only

5. Present SD1 and SD2 in random rotation.

(NOTE #2: There are some subtle intermediate steps that may be employed, but

these are the key components to the discrimination learning procedure)

Example Program: Receptive Object Labeling

SD: " Touch/find/give me/show me (object label). "

R: Finding the correct object from a field of any number of objects.

Field Size = two objects, cup on left side of table, shoe on right side of

table/step 5 in discrimination learning.

(NOTE#3: The following example is shows how win-stay-lose-shift (WSLS) occurs

when you use " No " to indicate an incorrect response. WSLS does not only

occur when you use " no " for feedback it may also occur when transfer trials are

employed, the difference being that it is subtle when transferring from prompted

responses to independent. No matter what prompting strategy you employ, if

the child is engaging in WSLS, your prompting is not working to teach the

response you are looking to teach)

Instructor: " Find the cup " (which is on the left side of the table)

Child: Selects the shoe that is on the right side of the table.

Inst. : " Nope (slight pause) find the cup "

Here the child selected the incorrect object, did not receive reinforcement

therefore he " lost " , so instead of focusing in on the OBJECTS, he thinks that

that POSITION on the table is not where he receives reinforcement, so the next

time he responds, he will SHIFT to the left side of the table.

Child: Selects the cup.

Instructor: " WOW! that was super, there ya' go!! " Hugs, more praise, and

tickles ensue.

(SEE! The child WON!, so now what will happen, he will select this position

again, because he thinks the POSITION on the table is what accesses

reinforcement, not the item.)

Instructor: Does not shift the field or add neutral distracters, and then

says, " Find the shoe. "

Child: Selects the cup.

Instructor: " Nope, find the shoe. "

Can you see how this can become a pattern? I've seen this occur in programs

where " No " is employed as feedback, in programs where " No " is not employed as

feedback but is not an " errorless " program, as well as programs that employ

errorless teaching. The reason why this would occur in errorless in teaching is

that at some point, the child has to be allowed to respond independently, if

the child errors or is about to error and is then prompted, therefore WSLS is

extremely difficult to identify.

Even though I have a visceral reaction to the term " no-no prompt " , let me say

that if this procedure is employed, WSLS is very likely to occur, but should

be easily recognized and then a different prompting and feedback procedure may

need to be employed.

I hope this helps clarify what WSLS is.

-Ray Cepeda