Kudos! ;-)

Doan

On Wed, 13 Dec 2006, Nathan A. Barclay wrote:

For anyone who hasn't been following it, the thread "Teenagers faced with
spankings" turned to a discussion of a study by Dr. Dennis Embry regarding
spankings and children entering the street. A question arose about a
quotation from Dr. Embry, so after Kane and I couldn't find an
interpretation we could both accept as valid, I tracked down an email
address for Dr. Embry and asked if he would be willing to clarify.

In the ensuing conversation, he provided a good bit of useful information
that clarifies his views on the problem of spanking and reprimands sometimes
causing children to enter the street more often. I asked his permission to
post the messages here, and he gave his consent.

---
My first message to Dr. Embry:

Are you the same Dr. Embry who worked on the Safe Playing program to reduce
traffic accidents due to children's entering the street? If so, I have a
couple questions I'd like to ask about a letter you wrote to Children
Magazine a number of years ago. I hate to bother you with this, but I've
seen the letter used to bolster claims that I'm extremely dubious about. So
if you're willing to take the time to offer some clarification, I'll greatly
appreciate it.

In your letter, you made the claim that spanking, scolding, reprimanding and
nagging increase the rate of street entry. Were you trying to say that this
is what normally results from parents' doing such things, or just that such
unintended results sometimes occur? And if you were saying that increased
street entry is the normal result, was that belief based on scientifically
valid research? Or could it have been a result of skewed perceptions
because children who keep entering the street even though they keep getting
in trouble for it are highly visible, while situations where children try
not to enter the street because they don't want to get in trouble are far
less visible?

------
Dr. Embry's first reply:

Hello Mr. Barclay,

Research and science are mean mistresses. I assume you are willing to have
whatever hypothesis you might have proved wrong? That is how I approach
things, as a more than passing good researcher.

I assume that you might be able to entertain that spanking could be good for
some kids, not good for some kids, and even harmful. Or, it could be
completely ineffective, neither good nor bad. If you cannot entertain that
results could be mixed or negative, you will not like any science that could
be completed. There are people who still insist that the world is flat, but
I don't think you are in that category if you are willing to write me.

I see that my work from 25 years ago is continuing to create all sorts of
crazy making behavior on various list serves. This amuses me.

Here is the nutshell

I did not set out to prove anything about spanking. That's a fact. So folks
can take a chill pill. In fact, I had recommended in my column and radio
show that it was OK for running into the street.

We did really good science, and you need to know that I am an A-list
behavioral scientist.

Basically, we discovered that kids who we would consider to be conduct
problems or oppositionally defiant (today) increased their rate of street
entry after spanking, nagging, scolding, etc.. Today, this no surprise in
developmental science. All this is behavioral psych 101. This is what we
call "accidental attention to negative behavior." That is, it functioned as
a reinforcer, not as punishment.

And no, these observations were standardized, with two or more observers. I
am not clear what your question is about observations otherwise. We
separated kids who were "high rate" versus "low rate." The high rate ones
were most interesting; the low rate kids were rarely bad, and responding
quickly to the interventions.

Let me know how I can help clarify.

------
My second message to Dr. Embry

Thanks for your reply! Would I be correct in interpreting your reply as
saying that you think only a relatively small (but still certainly
important) minority of children react to being spanked for entering the
street by entering the street more often? If so, that is very much in line
with what I expected would be the case.

By the way, I'm impressed with what I've seen so far about your Safe Playing
program, and I'm certainly not trying to denigrate its value. But the
question of how many children react to being spanked by misbehaving more
often has ramifications regarding a much wider range of issues.

------
Dr. Embry's second reply:

The issues raised in my original study and subsequently in the whole
behavioral literature on the eitiology of conduct disorders is very
significant, and most of the missives I have seen in the various list serves
(I have serious trouble trying to go through the Google one for some
technical reason that I don't understand).

We had about twenty kids in the precision oriented observational study,
using a multiple baseline. Today there are ways to calculate the
equivalent sample size and effect sizes, compared to a randomized control
study. The effect size of the safe playing study would be very, very large
because of the clarity of the repeated measures. Group designs only
estimate standard error: single subject designs directly measure standard
error. It would have been nearly impossible to have detected the fact that
spanking, scolding and reprimands served as "accidental attention to
dangerous behavior" except by a repeated measures, with 10-second coding.
That said, about a third of the kids had this effect, and they were the ones
that people often want to spank; that is, because these were the kids,
post-hoc, that would likely meet the definition of oppositionally defiant in
today's vernacular of the DSM-IV. The prevalence of this DSM-IV diagnoses
are clearly rising for a whole lot of reasons that have nothing to do with
parenting, yet parenting/teacher behavior can seriously worse the biological
and socially induced predispositions.

For these kids, spanking, etc. did not meet the operant definition of a
punisher; rather, it met the definition of reinforcement. This is whole
consistent with the long-term, precision studies of the etiology of
multi-problem kids (see the book by Anthony Biglan et al. Helping
Adolescents at At Risk, from Guilford Press). Dr. Biglan is my close
colleague and the president of the society for prevention research. Dr.
Biglan's synthesis book does a nice job of reviewing the cycle of coercion
work of people like Gerry Patterson and colleagues, which has been
replicated by other investigators. It is very parsimonious, and fits both
behavioral and evolutionary theory.

By the way, it is important to note gender effects. Boys are more prone to
have oppositional defiance and conduct disorders, and it is my opinion that
this is because of evolutionary pressure. About 1/3 of boys were killed in
neolithic societies as a result of tribal and clan wars; that is the
prevailing new authority or counter authority. Oppositional defiance and
conduct disorders seem to confer a sectionist advantage. Parental, teacher
and other coercion of children disposed to these traits (which now are known
to have polygenic mechanisms, not known even 10 years ago) clearly elicits
and then solidifies such a trajectory, causing immense social and personal
costs.

Folks might like to see the other papers I have published on what prevention
science is suggesting might work better to prevent these issues, based on
the continued unfolding of child development science.

Can you direct me to these list-serves that have all this going on? I would
like to read them more directly.

-----
My third message to Dr. Embry:

Thanks again. The discussion I'm involved in is on the Usenet newsgroup
alt.parenting.spanking. The discussion thread's title is "Teenagers faced
with spankings," and no one ever thought to rename the thread when it very
quickly drifted off onto other subjects. The discussion there is actually a
very small one, with just me and one other person engaged in anything
resembling real debate.

That discussion led me to look for more information on the Web. The first
Google hit using the search terms "dennis embry street children" comes up
with the page http://www.neverhitachild.org/embry.html which precedes the
text of your letter with the caption, "'Spanking... increases the rate of
street entries by children', wrote Dr. Dennis Embry in a letter to Children
Magazine." The third Google hit on those terms points to a closely related
page, http://www.geocities.com/cddugan/SpkEmbry.html which also tries to use
your letter to support a blanket denunciation of spanking. (That becomes
especially clear looking at the web site that the page is a part of.)
Unfortunately, if the study itself is available on the Web, or even good,
unbiased information about it, that information was either far enough down
in the search results or hard enough to recognize from what Google returned
that I missed it with the combinations of search terms I tried.

I have two more questions, if you don't mind. First, what underlying
population were the "about twenty" kids drawn from? That has important
ramifications regarding what the "about a third" you refer to is about a
third of, and I don't currently have a copy of the study to check.

Also, my expectation would be that in borderline cases, children could make
a game out of resisting minor adverse consequences in order to get attention
but not view the game as worth the cost if the consequences are more
serious, or could resist minor adverse consequences because they don't like
being told what to do but regard the cost of resisting more serious
consequences as too high. Does that fit with your understanding of
children's behavior? (I'm definitely aware that it can be dangerous to
repeatedly escalate punishments that aren't working, but when I see a danger
in one extreme, I try to look for the possibility that the truth is
somewhere in the middle rather than rushing to the opposite extreme.)

------
Dr. Embry's third reply interspersed his response with what I'd just written
(above). In order to post plain text, I'm reformatting to put NB in front
of the lines of material he quoted from me. Also, he included a graph
which, unfortunately, can't be included in a post to a text newsgroup

Dr. Embry's third reply (reformatted):

Thank you for this info.? I will answer the questions further below.? I
didn't have the study available, as I was on a plane (and it is on paper
only, not PDF, etc. having been done in 1979-1981.? I could not remember all
the details.? I am going to have to scan this damn thing and put it up, or
resubmit it to JABA or some such, as it is one of the few actual
experimental studies on pedestrian safety in the world.??

On Dec 13, 2006, at 3:01 PM, Nathan A. Barclay wrote:

NB Thanks again.? The discussion I'm involved in is on the Usenet
NB newsgroup alt.parenting.spanking.??The discussion thread's
NB title?is? "Teenagers faced with spankings," and no one ever
NB thought to rename?the thread?when it very quickly drifted
NB off onto other subjects.? The discussion there is actually a
NB very small one, with just me and one other person engaged
NB in anything resembling real debate.
NB?
NB That discussion led me to look for more information on the
NB Web.? The first Google hit using the search terms "dennis
NB embry street children" comes up with the page
NB http://www.neverhitachild.org/embry.html which precedes
NB the text of your letter with the caption,?"'Spanking... increases
NB the rate of street entries by children', wrote Dr. Dennis Embry in
NB a letter to Children Magazine."? The third Google hit on those
NB terms points to a closely related page,
NB http://www.geocities.com/cddugan/SpkEmbry.html?which also
NB tries to use your letter to support a blanket denunciation of
NB spanking.? (That becomes especially clear looking at the web
NB site that the page is a part of.)? Unfortunately, if the study itself
NB is available on the Web, or even good, unbiased information
NB about it,?that information?was either far enough down in the
NB search results or hard enough to recognize from what Google
NB returned?that I missed it with the combinations of search terms I tried.
NB?
NB I have two?more?questions, if you don't mind.??First, what
NB underlying population were the "about twenty" kids drawn from??
NB That has important ramifications regarding what the "about a third"
NB you refer to is about a third of, and I don't currently have a copy
NB of the study to check.

There were a total of 33 preschool-age children in the study, all but three
enrolled in the university affiliated preschool.? The school had a mix of
children and parents, including normative and high-risk kids.? I have a
table of special characteristics, if known.? Eleven of the 33 had some
"condition, such as aggression, language delays, etc.? Out of the 33
children/families, 13 of those were intensively observed.? One of the
observed kids (S4) clearly had a "label" of what we might now call
oppositional defiance.?

Here is the graph of the "high rate entry" kids.

[Graph lost in conversion to text format]

These kids needed both the reinforcement/self-modeling AND sit-and-watch, a
variant of Time Out (but emphasizes re-engaging in the positive behavior
ASAP.? Sit and Watch interrupts the negative reinforcement for bad behavior.

The low rate kids responded to reinforcement and self-modeling alone.

NB Also, my expectation would be that in borderline cases, children
NB could make a game out of resisting minor adverse consequences
NB in order to get attention?but not view the game as worth the cost
NB if the consequences are more serious, or could resist minor adverse
NB consequences because they don't like being told what to do but
NB regard the cost of resisting more serious consequences as too high
NB.? Does that fit with your understanding of children's behavior??
NB (I'm definitely aware that it can be?dangerous to repeatedly escalate
NB punishments that aren't working, but when I see a danger in one
NB extreme, I try to look for the possibility that the truth is somewhere
NB in the middle rather than rushing to the opposite extreme.)

We saw kids get their butts hit pretty smartly in baseline, then go into the
street AGAIN within a few seconds or minutes, showing the mathematical
relationship of a reinforcer.? That only was true for the high-rate kids,
though.??

This is what caused my jaw to drop, observing the temporal sequence of both
the topography and function of reinforcer.? One sees this in micro-coding of
regular, daily parenting in the studies such as Hill Walker's and Gerry
Patterson's of highly deviant kids and families.? Those kids tend to get
nuked, but I never expected this in the context of dangerous behavior.? I
should scan the pages on the time relationships.

We had one child and parent that showed no behavior change at all, except
for the brief modeling effect (that we saw in the earlier study) S13.? This
child's parent was one of the "worst offenders" of negative attention, and
never did any positive attention that we observed.? Children with such a
serious imbalance are very high risk for developmental pathologies.? This
would be the type of parent who alternates between very permissive and
highly punitive.?

If I were to make a thumbnail of the findings (and informed by other
research), spanking kids who rowdy attention seekers (mostly boys) as young
children is likely to backfire and increase deviant behavior. This is a
functional, empirical assessment, not a moral or religious one. This effect
is almost certain if the positive attention for the child's behavior is
below accidental attention to negative.? Very nice longitudinal data on
this.? It is the frequent reliance rather than very, very rare reliance on
spanking that seems to have adverse effects. (All this follows a very nice
mathematical law, called the Matching Law.)

Parenthetically, the Safe Playing study was being done concurrent to our
other work at the university of kansas parenting program, where we did
direct observations of families at home using very precise observational
codes every 10-seconds, with independent observers.? About 85% of the sample
had open case files with child protective services, and our observers
routinely witnessed what can only be described as physical hitting
(spanking, slapping, pinching, etc.) many times per hour in about 80% of the
families (85% x 80% = 68%).? We never observed such things in the normative
families.? These families had very, very low rates of positive attention,
very high rates of negative attention and the children were singularly
awful.??

We taught the parents to engage in a behavior change project and keep data.?
We also coached the parents at home and in public settings to increase
positive attention, reduce negative attention, and use various consequences
such as Time Out, put-throughs, over-correction, etc. (Time is not too
effective for under 3-year olds). These families became very skilled, and
their results maintained 18-24 months, with no repeat fillings with child
protective services.? ?High end families cannot learn to do these skills
without home-based coaching.? Only two-parent, intact families, with
low-stress and normal kids could take general parenting skills and apply
them at home without coaching.?

You are correct about the extremes. It is a bit like the Iraqi folks
fighting over their sects.? Not functional.

I prefer to take a clear look at the functional relationships and build up
from that. I hope this conversation helps.
.