Functional Analysis pt. 3

Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Speaker 1 (00:03):
You're listening to why we do what we do? Welcome
to why we do what we do? I am your
exploring all conditions, Host Abraham.

Speaker 2 (00:21):
And I'm you're isolating those conditions.

Speaker 1 (00:23):
Oh, Shane, we are a psychology podcast. We talk about
the things that humans and non human animals do. For
those of you joining us for the very first time,
then that was probably a weird thing the way that
we introduced ourselves, but it will make more sense as
we get into this. We are on the third part
of a multi part series of episodes discussing something that

(00:44):
is deep in the weeds of behavior analysis practice and
literature and sort of understanding called functional analyzes. This is
the third one, and so if you haven't heard the
last couple that we have done, then you recommend checking
those out for some background and conceptual pieces that will
be necessary. But we'll also do our best to recap

(01:06):
those going into this. And if you're a returning person
then welcome back. Then that recap will be because you
haven't heard those at least a week or a couple
of weeks, and you're getting caught up too, and this
way you recall what we're doing as we get into it.
But however you got here, we appreciate it. And if
you do find that you would like to support us
after all of this, then you can do that by
leaving us a rating and a review like subscribe, tell

(01:29):
a friend, join us on Patreon. I'll talk more about
those things at the end of this discussion. But it
is August twelfth, as you may or may not know,
depending on when you're listening to this, it might not
be August twelfth anymore. But when that's publishes, it's August twelfth,
which means, of course, Happy Baseball Fans Day. That's me.
So they they put oscillating fans in the baseball field

(01:50):
and they blow away.

Speaker 2 (01:51):
Yeah, yeah, they have like a good windy day depending
on where the wind is blowing. That's a baseball field.
Batters really love the wind blowing out the center field,
you know.

Speaker 1 (02:00):
Yeah for sure.

Speaker 2 (02:01):
Yeah, pictures not so much. It is ibm PC Day.

Speaker 1 (02:05):
That's a lot of initials.

Speaker 2 (02:07):
M hm.

Speaker 1 (02:07):
It is International Youth Day.

Speaker 2 (02:10):
Except when we were younger, we started a band or
we were gonna we were trying to name our band
and we were gonna call it youth in Asia.

Speaker 1 (02:16):
Yeah, that is a band.

Speaker 2 (02:18):
That is a band and it was we didn't know that,
but at the of your like, we're so clever. It's
a play on words. So here's the International Youth. Yeah.

Speaker 1 (02:26):
I thought you're gonna call yourselves international youth. I was like, man,
that would have been quite the coincidence.

Speaker 2 (02:30):
Yeah, yeah, I mean close enough. I guess it's Julian Fry,
Julienne Fry's Day, whoever that.

Speaker 1 (02:37):
Is, who's frying Julian. It's a middle child day, it is.

Speaker 2 (02:46):
It's also National Sewing Machine.

Speaker 1 (02:47):
Day Handy Devices. It is truck driver Day. Oh, we
need those people. We do desperately, the bone of the economy.

Speaker 2 (02:57):
Uh huh uh huh. It is Vinyl Record Day.

Speaker 1 (03:00):
A nice, I dig it. It is World Calligraphy Day.
In case you don't want to read what it is
you're presented with in text form.

Speaker 2 (03:09):
Yeah, yeah, for sure. It's World Elevant Day. I love
those things.

Speaker 1 (03:12):
Yeah, selfants are amazing. It is Elvis Week.

Speaker 2 (03:16):
It is. It is also National Bargain Hunting Week.

Speaker 1 (03:20):
Mmmm okay for our thrifty folks out there. It is
National resurrect Romance Week because before this it was.

Speaker 2 (03:27):
Dead, yes, and now we get zombie romance.

Speaker 1 (03:31):
Nice.

Speaker 2 (03:31):
And it is Save plus sound Week.

Speaker 1 (03:34):
Did you ever see the movie Warm Bodies.

Speaker 2 (03:36):
No, but I heard it was really good. It's Nicholas Holt, right,
like yeah.

Speaker 1 (03:40):
Yeah, yeah, and the whole it is zombie romance, so
it was fun. I recommend checking it out. Also, just
know what you're getting into, is you go into a
movie that is a romance with zombies.

Speaker 2 (03:51):
So it's very particular.

Speaker 1 (03:54):
Yeah yeah, not maybe for all audiences, but I thought
it was very fun, Like it is a adjacent to
horror rom com It's somewhere there. Yeah, genre blending and bending,
that's always fun. Anyway, we're not here to talk about
zombie movies or even necessarily just the holidays. We're talking

(04:16):
about functional analyzes, and just to recap what we've discussed
really quickly so far in case you have never heard
of these and what we're talking about. This is essentially
an assessment strategy very commonly employed by behavior analysts in
which they create some conditions to observe what could be
challenging behaviors, most often with individuals who have limited communication capacity,

(04:42):
and when those individuals have some types of behaviors that
are very challenging that could be harmed to themselves or others.
Or other like related type things. It's like something is severe,
let's say, in the way that they're behaving, and we're
trying to figure out why are they doing this thing
so that we can potentially address how do we get

(05:03):
their needs met without them having to do these destructive
sorts of behaviors.

Speaker 2 (05:08):
Yeah, and we also covered you know, as we kind
of are talking about this, we talked about the arrangements
and different variations of these too, because there's lots of
different ways that we do this. So like we set
up these condidencies, we set up the way we these
tests essentially, and then we do them in a lot
of different ways because we really ultimately want to find
out this concern is happening, right, this behavior is happening.

(05:29):
We want to find out a causal relationship, Like that's
the goal. It's not a correlational thing. And we have
to do this in different ways because a there's no
such thing as a one size fits all type of
assessment and so so I think that's really important to
note too, is like there's some unique flexibility inside of
this what could be perceived as a rigorous test type

(05:52):
of assessment. So there's some really cool stuff in there
as far as that goes.

Speaker 1 (05:56):
Yeah, we impact a bit of the history of how
this developed, starting before it was actually an experimental sort
of thing, with it being more of a conceptual exercise,
and they called it a functional analysis, and then in
nineteen eighty two the first study where it turned into
an experimental functional analysis. And essentially the idea here is
sometimes compared to an allergy test. It's let's test a

(06:19):
bunch of different situations and see which ones are the
ones where we see that challenging behavior occur, and then
that tells us what is provoking it and what is
maintaining it.

Speaker 2 (06:30):
And this is how we figure out you're allergic to math.

Speaker 1 (06:33):
Yeah, exactly. That's also funny inside joke because one of
the original studies had to do with how difficult certain
academic tasks were. But anyway, the whole point then is
that once we understand that, we can use that information
to help guide Okay, this is what this person needs
in this situation. Now we can teach them to communicate
what they need rather than do this thing that is

(06:54):
harmful to themselves or others are both, And then where
we went from there is essentially various iterations of people
exploring what works, how do we change this, what can
we make different about it? That sort of thing.

Speaker 2 (07:07):
Yeah, and so it kind of led us up to
a little bit of where we are at today, which
is is kind of understanding how the actual assessment works. Right, So,
like we're looking at like we've established these conditions that
we test, we've isolated variables, and here we are kind
of looking at I think that what we've found is

(07:28):
that the isolation piece is kind of a core feature
of this. Like I and when I say isolation, I
mean isolating particular like related variables to whatever the conditions are. Yeah,
we're going to see that across every type of assessment,
but we're at the stage now where we're kind of
like maybe even moving away from that a little bit.
And we'll kind of talk about that today, I think
when we talk about the synthesized types of tests. But

(07:50):
I mean, it's really come such a long way, even
in our years of working in the field, because I
know that when we first started, some of these variations
didn't even exist.

Speaker 1 (07:58):
Yeah. Yeah, and so we got into essentially describing what
the conditions are. For the most part, there's a condition
that tests for positive reinforcement, a condition that tests for
negative reinforcement, a condition that tests for self stimulatory or
sensory reinforcement, and then all of that's tested against the
control condition where it should be no motivation to do
these behaviors at all. And then there was latency based fays.

(08:22):
There are trial based defays their natural environments or pop
up or burst contingency fas. An FA, by the way,
stands for functional analysis, in case I hadn't said that already,
and so we'll probably call it sometimes a functional analysis
and sometimes an FA. But that's more or less where
we left off. And that kind of brings us up
to the discussion today, where we're going to start by

(08:42):
getting into another way to sort of arrange these conditions
to test for possible functions, if you will, Yes.

Speaker 2 (08:50):
And I'm so excited about this because I think that
they're so I do think this is such a cool
way to discover relationships about behavior, right, Like it's so interesting,
and like I would say, as somebody who has done
several of these, it's so cool to watch. It's cool
to see like some patterns e merge as you're doing
these types of assessments.

Speaker 1 (09:09):
You know, yeah, you really can see when you're observing
and participating in conducting these you get to see it
with your own eyes, the sort of differentiation that can
occur between those conditions where it becomes fairly clear, like, look,
this is we're seeing something a pattern emerge. I think,
as you said, a trend. So yeah, all right, perfect,
I think that gets us caught up to where we're

(09:30):
at and so getting into this. Then another way that
they might specifically try and assess and determine what's going
on for a particular individual is a strategy for presenting
these fays where what they want to do is directly
compare just one single condition, one set of variables, or
like one variable basically and then test against that control condition.

(09:52):
This is called a pair wise analysis pai R not
par yeah, pair wise, where you have basically just those
two things that you're looking looking at. And this is
most often implemented historically when a functional analysis was not
yielding very clear outcomes and there was sometimes variability across conditions,
sometimes there was very little behavior occurring across the conditions,

(10:15):
and a way to determine if there was a functional
variable related to the contingency to the outcomes that the
individuals trying to get to is to expose the individual
to that condition only just those variables, and then alternate
that with the control conditions. See like, if in this
one condition that we're repeatedly going to test, alternated with
this control condition where there should be no instances of

(10:38):
this challenging behavior occur, can we there see a differentiation
happen in the pattern where it's clear that it happens
over here and it doesn't happen over there.

Speaker 2 (10:47):
Right, And so like this one's really interesting because I
think historically what we see with the other types of
arrangements is it'll go, you know, let's number all the
conditions one, two, three four, right, so it might like
what we kind of set up the test, it'll go
one two one, you know, one two three four one
two three four, and then like we'll kind of alternate
between those. What's interesting about a pair wise is it

(11:09):
to go one two, one two one two for a
few data points, and then we'll test another condition and
go one three, one three, one three with one being
the control and just kind of to see if there's
a trend like isolated instead of having that alternation between
one and two and three and four and kind of
doing that over and over and over again. So like
that's kind of what it might look like in practice.

(11:30):
I mean, unless unless I'm completely off base there, but
it's just to kind of like give people exactly what
I like to me, I think it's important to see
what that looks like.

Speaker 1 (11:36):
No, you're exactly right. That was actually a really that
was I think a really nice simplified way to describe
exactly what the pair wise analysis is between like actual
content and ads. We're back. Yeah, So just as you
just said, very well, very well put that, it's basically like, yeah,

(11:58):
let's j just do yeah one maybe being control conditions
and two being the variable and then one two, one,
two and going back and forth between those, and then like,
depending on how that goes, you might then test another one.
And this will be important later when we get to
some further innovations in functional analysis implementation strategy, because that
idea of like let's just test a thing against a

(12:20):
control will come back again later. But yeah, in general,
these were applied after a standard functional analysis had been implemented,
Like they'd tried to just go through the normal procedures,
but it failed to produce any clear outcomes. Like we
were running these conditions, but we don't see a pattern
emerging now.

Speaker 2 (12:38):
And if we.

Speaker 1 (12:39):
Described in a previous discussion that this could mean this
quote unquote automatic function, and we described that in the
very first episode in the series so far, and not
to dwell on the philosophical issues that might be raised
by calling it quote unquote automatic, but for those of
you who know, you understand, and if not, we did
discuss it a little bit back then. But anyway, it
could also mean that the reason that we don't see

(13:01):
any clear patterns emerge in the functional analysis is because
there was insufficient discriminability, meaning it was just not clear
to the individual in question what the difference was between
the conditions they were supposed to be experiencing. And so
if we simplify the arrangement to just one potentially evocative
condition against that sort of control or neutral condition, that

(13:23):
might help illustrate the possible reinforce or the possible provocative
situation and corresponding outcome that the individual's trying to get.
And it's like for them, it was just like, for
some reason they do this here, and then some reason
they do this here, and they don't really understand what's happening,
because most of the time these are individuals very low
developmentto their language, and so we can't just ask them

(13:44):
or explain to them what's going on. And so it's
simplifying it to just this is one thing, and this
is the sort of neutral nothing should happen thing, and
just make that distinction really clear could help provide the
circumstances under which it does become more clear to the
individual and to the than the clinicians what's going on.

Speaker 2 (14:02):
Yeah. Absolutely, So for example, let's say there's a functional
analysis that had variable patterns, and what we mean by
that is like just spiky all over the place, kind
of looking like what a Rechdor scale might look like,
just like bounce it all up some high, some low.
Yeah yeah, yeah, it's wild. So that's variable patterns with
no discernible trends across those conditions. So like you can't
figure out what's happening. It looks like every all the

(14:24):
behaviors is kind of all over the place. The clinician
might try presenting only the very specific attention condition identified
in the interview and alternate that condition with the control
play or neutral condition like we mentioned. So that's the
one two, one two, one two versus one two, one three,
one four, and so on and so forth. And if
there is a clear separation in the data, it might
suggest that the form of attention, or at least one

(14:46):
of the forms of attention that are occurring, or some
variable there is maybe one of the maintaining variables, if
not the maintaining variable. That separation really makes the difference there.

Speaker 1 (14:56):
Yeah, just being able to see like, Okay, when we
have just this one, it's clear that we see that
those challenging behaviors existing, but in this other one we don't.
So that helps us then differentiate. Okay, now we see
some kind of pattern we can then use. If the
data do not show clear separation, the clinician might try
a different pair wise condition against that neutral control condition,

(15:17):
just as you were saying before the whole example of like,
let's try this specific form of attention and do that
against alternating with that neutral condition, back and forth, back
and forth, back and forth. Still don't see anything, all right,
Let's try a different type of attention condition and alternate
that back and forth. Or let's try a different type
of escape condition and alternate that back and forth. And
there have been many studies summarizing data from functional analyzes

(15:39):
from dozens of experiments, and that's honestly underselling it. Probably
hundreds of experiments at this point showing that the pair
wise arrangement can be really helpful and identifying a function
in which the function of the behavior, like the outcome
that individuals seeking, was not readily detectable in that standard arrangement.

(16:00):
And so essentially you look at the published literature. There
have been many, many, many studies at this point where
they summarize dozens of cases at a time and sometimes
over one hundred, and then looked at like, how often
what was the pattern we saw, how often did it
emerge that we saw this, How often did it emerge
that we saw this? How often was it the case
that we didn't see anything clear.

Speaker 2 (16:19):
In the data?

Speaker 1 (16:20):
And then how often in those circumstances was it the
case that a pairwise assessment did yield some clear trends
or patterns of the data. And there's quite a few
that show that they did.

Speaker 2 (16:30):
Absolutely Now and many of you who have ever worked
in a clinical direct service delivery capacity, you know you're
working directly with clients, you're working family homes, you're working
at schools, whatever might be. You might have had the
thought while we're talking through FAS that while this assessment
seems ideal and it seems like a great way to
develop relevant interventions, it might be difficult or nearly impossible

(16:50):
to implement in day to day operations in many places.
And I've also heard that funding sources will not pay
for additional time to do functional analysies unless it's a
private pay space. But that's also a barrier that comes
up too, is like not only is there like maybe
does it seem difficult to do, but you might not
be able to build an insurance company for it.

Speaker 1 (17:12):
That's a great point too. Yeah, there are variety of
reasons people have cited as to why they cannot feasibly
implement a functional analysis, and in response to that, first,
some scholars have taken umbrage with that argument, and we'll
outline one of the most seminal of those in just
a moment. But second, there was another innovation was that

(17:34):
perhaps the behavior function could be assessed simply by observing
a natural environment in which the behavior could occur, rather
than creating that evocative situation in sort of an experimental
artificial environment. They're like, well, maybe what we could do
is we could just watch this person and their sort
of normal day to day and see if we can

(17:57):
determine just observing them that way why those behaviors occur.
Or maybe something that's even easier than that that could
be sufficient enough to determine what the function of that
challenge and behavior might be, is to just use a
sort of standardized questionnaire with the relevant stakeholders in that
person's world who might have enough experience and insight into

(18:17):
those behaviors that we could glean the function of their
behavior without direct observation and without that experimentation. So essentially
the direction we're going right now is that people sort
of argue, Okay, well a functional analysis seems great, but
we just can't do it. So what can we do instead?
And the two options are maybe we directly observe this
behavior in its natural environment. Maybe we just ask the

(18:40):
people who are around what is going on the way
that they understand it.

Speaker 2 (18:44):
But there is an excellent article, and I love that
it comes from this person because of their current work. Yeah, yeah,
this was actually like a big schism in our field
when this person went off to do their other things
and actually from what I understand, cause some headbutting between
I want to in this person. Yeah, even though this
person was a student of iuaas so Greg Hanley and

(19:05):
twenty twelve in the Journal of Behare Analysis in Practice
passionately pointed out some of the arguments against fays and
provided solutions to those arguments. So, like, this is one
of those articles. It's just kind of like, oh, here
are nine myths that people. It's a really good, like
listical type of article. It's amazing. I mean, it's genuinely
done so well because it's like, here's an argument, here

(19:27):
are some examples in our literature. Here's an argument. Here
are some examples of our literature that debunk that. So
it's basically like a myth busting article, which is so cool.
So yeah, the first argument is that people will say, oh,
functional analyzes take too long, and he offered five minute
sessions or trial based sessions or latency based fays which
are much quicker. So like, there is a solution to
that it takes too long argument, because there are variations

(19:50):
that do reduce the amount of time it takes to
assess the behavior.

Speaker 1 (19:54):
And actually for all of these like we're only going
to list a few that he gave. There are quite
a bit more that we're not going to get into,
but I do. I did put the article in our
links and references for people who want to go check
it out because it is really really well written and
has some really useful content. So he does provide like, Okay,
if this is your problem, here's a potential solution. He

(20:14):
also provides like, here's why this is also not really
a problem, right, But anyway, we'll get into that. So
the next one, that argument is that it's too complex,
it's too hard to train staff, it's too hard to
implement with integrity, et cetera. And so he suggests that
one way to address that is using those pairwise assessments
using one of the most relevant condition identified during the

(20:35):
open ended interview portion of it. That way, you're only
then doing more streamlined implementation strategy.

Speaker 2 (20:42):
Yeah, one of the arguments that comes up often is
that the functional analyses are too dangerous, it's too dangerous
to evoke behaviors, and like, there is some truth to that.
But what he did was he offered these solutions around
conducting assessments in safe places, clear condition signaling, brief sessions,
and reinforcing pre behaviors only if you can find a
reliable precursor. I think he also in that argument he

(21:04):
talked about even identifying an arbitrary behavior that is completely
unrelated to the problem behavior and doing an FA on
like some like an arbitrary unrelated behavior, and you could
still derive a function from that, which I thought was
kind of interesting. And he uses I think that argument
later to get into some of the stuff that he's
doing today which has demonstrated effect. But as somebody who
has done fas for dangerous behaviors, I can attest that

(21:27):
like if you do precursor fays, or if you do
latency based that phays like ones where you do not
allow the behavior to occur, you could still get the
function out of it. I had to do an FA
on a skin picking behavior that was causing wounds in
the person. I had to do an FA on a
rectal digging behavior that resulted it would go from rectal
digging to physical aggression pretty quickly. And then we had
to do one on a self injurious behavior where a

(21:48):
learner had something like nine thousand, roughly nine thousand incidents
of SIB per month. Wow, And we were able to
do fas with all three of those and get good
effect sizes.

Speaker 1 (21:58):
Yeah, and I don't recall now if in previous discussions
we mentioned the precursor behaviors. But essentially the idea here
is that in the class of behaviors that are all
around sort of in the same vein or the same
topography class, if you will, of the challenging behavior of question,
a lot of times they sort of ramp up to

(22:19):
the high magnitude challenging behavior. And so those behaviors that
sort of signal to you like, oh, they're getting worked
up now, you can actually reinforce those rather than letting
them get all the way to full escalation. Right, And
that because they're all sort of part of the same momentum,
if you will, you can sort of stop it early,

(22:40):
still reinforce that and probably detect a meaningful function of
the behavior. And these are actually super handy. I've done
a few of these as well, and the research is
really clear, Like that's pretty actually as usually that's often
very effective for identifying what the reinforcer is that is
relevant in this particular situation without allowing it to escalate
too far. Yeah, that's definitely one to consider as well.

Speaker 2 (23:02):
Yeah, I think precursor behaviors are pretty accurate predictors too.
Like there's some Tim Volmer does some really cool research
on precursor behaviors and like some specific formulations that will
predict them, and like you and I have kind talked
about this before where we have talked about precursor behaviors
where like sometimes you'll start hearing my voice pitch up
a little bit, like I'm almost like preparing to deliver

(23:23):
some news that somebody doesn't want to hear, and you
can kind of like figure out that it's a little
bit of a different cadence than my normal podcasting voice,
and that usually means that an AD's coming.

Speaker 1 (23:39):
So we've been discussing some of the arguments that occurred
inside this article that was essentially trying to point out,
like these arguments against functional analyzes are there are ways
to address the concerns and arguments that people have here.
There's like it takes too long, it's too complex, it's
too dangerous. He also argued that while functional analyzes do

(24:00):
take time, is better time spent than to spend time
having not done an FA to then create an intervention
that is less likely to succeed and will ultimately contain
fewer reinforces for the individual than they would have been
able to identify if they had correctly found what the
behavior function was by having done the functional assessment itself.

(24:23):
And so he's sort of pointing out that like, Okay,
it does take time, but that's efficient because otherwise you're
going to waste a bunch of time doing stuff that
doesn't work very well, and that the individual in question
here is now also not having as enriched of an
experience as they could have otherwise had.

Speaker 2 (24:40):
Right. I've actually heard some will call and make that
same argument, which is like and it's it's ultimately an
ethics argument, right, like, yeah, is it more ethical to
avoid doing an FA because you're worried about problem behavior
but to waste time on interventions that are not effective?
Or is it more is it more ethical to like
do the assessment to identify the function and design more
effective intervention earlier and like that that's ultimately the ethics

(25:02):
question I think I land on, Like when we have
these conversations, it's like, yeah, yeah, let's not waste time.
Time is a commodity that we don't get back, right, Yeah,
So in an article, he provides an analogy to being
intubated in a hospital bed and extremely thirsty but unable
to communicate. All kinds of people are attempting to help
you and make you feel better, getting you a new pillow,
changing the TV, bringing you things that you like, but

(25:24):
because they don't assess what is you need, they just
provide things that are nice but don't actually solve your problem.
They're wasting time while you're still thirsty.

Speaker 1 (25:33):
Right, And so the whole situation there being that like, well,
if it's not what the individual actually needs, then like
it's all well and good that we're trying to do
nice things for them, but like it's not it's not
solving an issue. And that's that's the similar to the
ethics piece of like I mean that basically is it's
just an analogy that he created. But yeah, if you're
not actually addressing the behavior function, you're trying to essentially
brute force your way through this by just providing some

(25:55):
arbitrary something or other for something or other that's not
actually getting at with this person need in the situation
and therefore why they're engaging in that behavior in the
first place. And then, as I mentioned, before. In this article,
he provided a multi page table outlining some of those
arguments and concerns with conducting a functional analysis, and then
some possible solution to those problems. Again, that link will

(26:15):
be in our references here, but yeah, so that's essentially
a thing to check out if you're curious and learning
more about the sort of arguments that are made and
how he addresses all of those. He further took some
issue with the idea of these descriptive analyzes and these
observational assessments and the questionnaires, because we mentioned there's sort

(26:38):
of the there's the people who are saying like, well,
I can't do an FA. Handley says, yes, you can,
and here's how and why. But he also points out
in this paper and one of the other things that
people have said is like, well, if I can't do
an FA, then what I can do is just directly
observe the behavior in its natural environment and just record it,
take some data, and do an assessment that way. And

(26:59):
also maybe what I can do is implement a questionnaire
like I can just standardize format, interview the people who
are relevant and around this person who around these behaviors,
and see if they can provide some insight here. He
also pointed out that there are some critical flaws with
this that just do not do a good enough job
supplementing the functional analysis. But let's get into what those

(27:21):
consists of.

Speaker 2 (27:23):
Yeah, so let's laborate on the observational analyzes piece. So
these are commonly what is meant by a functional behavior
assessment or FBA. That's when we go in and we
do the observations, we do interviews, we come up with
a clear objective description of the target behavior of concern,
whatever that problem behavior is. We call that an operational definition,
and then we might hypothesize a potential function from this,
and that might include you know, things like ABC data

(27:44):
and whatnot that have given us some like patterns that
we have observed or the people have reported to us.
But either way, at this point, the clinician observes the
individual in some environment in which the behavior might occur,
and so we're observing when these events actually happen and
kind of letting the play out and capturing some data around,
like the variables that are contributing to those moments.

Speaker 1 (28:05):
So the person who is trying to do this functional
behavior assessment or FBA, what this might look like, is
they're going to try and go into the environment where
that behavior, the challenging behavior, takes place, and they're going
to note the full context whenever the behavior should occur.
What they're trying to identify here is what is the
circumstance that immediately preceded the target behavior, and they call

(28:26):
that the antecedent. What is the behavior itself? And then
what are any outcomes that are achieved by this behavior,
which is the reinforcer or the consequence? And then, as
you just mentioned, this is called an ABC analysis. A
for antecedent, B for behavior, C for consequence, and it
just flows really nicely. ABC analysis easy to remember. It's
a simple three term contingency assessment attempting to describe possible

(28:50):
triggers and reinforcing outcomes that may be facilitating the occurrence
of those behaviors the sound challenging behaviors or target behaviors.

Speaker 2 (28:58):
Yeah, and for people who don't understanding behaviorism, they might
hear that and think that it's a little reductive that
they're going like, oh, that's all behavior boils down to this.
There couldn't be more reasons for this, YadA ya YadA.
And as a behaviorist, I would also say it can
be reductive because it may not capture all the things
that it needs to capture. It does its job for
what it does. It captures information that you can use

(29:20):
to develop further hypotheses. But I'm going to challenge every
behavior analyst that listen to this that when you go
and you are doing a functional behavior assessment and you're
doing FBA's and all that, and if you rely heavily
on ABC data as the primary driver for your hypotheses
and that's like what your go to is, you may
be missing some information and it's not sufficient. It's just

(29:41):
not sufficient in like the sometimes it is that simple.
Don't get me wrong. Sometimes behavior is that simple. It
truly is. But there are other times where there are
just other variables of play that we often miss in
those types of analyzes. So I would just challenge everybody
to like think a little bit deeper about the way
that you do assessment with those.

Speaker 1 (29:58):
Yeah, we'll get into some of Handley's specific arguments against
depending too heavily on those descriptive assessments as well.

Speaker 2 (30:05):
Yeah, let's do that now. So, for instance, if the
clinician is observing the individual who tends to bang their
head on a desk in a classroom. It might be
the case that when the teacher is busy interacting with
other students and the individual in question is sitting by themselves,
they strike their head against the desk. The teacher rushes
over to stop the head banging from happening, and the
antecedent is the withdrawn or divided attention, like the attention

(30:26):
is not there, right, and the behavior is the head
strike to the desk. And then the consequence is the
teacher providing direct one on one attention whenever that happens.
And so this might be the reason the attention is
the primary reinforcer maintaining self injury to this individual in
that particular context.

Speaker 1 (30:43):
Yeah, just as an example, like antecedent attention gone behavior,
headbanging on the desk consequence, they get attention, right, And
so that ABC assessment there would tell you like, Okay,
this person wants attention, and that's so they do these
they act out in this way because when they do that,
they get the kind of attention they're looking for. Right
of note, and Handley points this out too, there are

(31:03):
some drawbacks to this simple direct observation strategy for one thing.
As Hanley points out, and this is definitely true, most
people in this individual's life are generally attempting to mitigate
the circumstance that produced those challenging behaviors. So there are
often few opportunities to actually observe that behavior happening. Right,

(31:26):
Like people have figured out, like, here's how to not
set them off. This person has spent a very long time,
the individual question here sort of teaching the social environment
how to not get them pissed off and not engage
in these behaviors. And those people have learned from that
and figured out like, let's try and avoid it. And
so that can actually result in some other complications when

(31:48):
trying to do an FBA, right.

Speaker 2 (31:50):
So FBAs can actually take a really long time because
there may be those few opportunities to observe. And I
think that's something that does happen. I mean I did
an FBA one time. It was five weeks, yeah, because
I was told that this person had problem behavior, and
it was five weeks of collecting data waiting for the
behavior to occur, and then ultimately the behavior didn't occur.
That the behavior interest didn't occur at all, Right, or

(32:11):
happened once. He was a forty five year old man
that said ass one time, and everybody was like, oh
oh no, So I was like, he doesn't need services.
But it was five weeks of observations, of data collection
of interviews and it yielded zero results. Yeah.

Speaker 1 (32:25):
Yeah, And that's a very long time as opposed to
like the cumulative twenty five minutes that might have been
spent in a functional analysis. And also, if your target
behavior is swearing, then pull your head out of your ass,
people like this is this not? Yeah, grow up, it's
not that big a deal. All right. Another drawback here
that is worth considering when you're looking at doing something

(32:47):
like a descriptive assessment or observational assessment of some sort
is that there are very very few I'm actually not
aware of any, but I have not dug into the
literature in a while, but there are very very few,
if any, peer reviewed studies that have shown that an
effective intervention was developed from one of these FBAs from
this observation alone. That doesn't mean that that doesn't happen,

(33:11):
Like I think that there are plenty of instances where
there are times that they do develop some effective interventions.
It is just very underrepresented in the published scientific literature
in the extent to which it is effective. So it's
not to say that it doesn't ever work, but the
science just right now seems to not really back that
up as a very valid strategy for conducting this assessment right.

Speaker 2 (33:34):
And at the end of the day, FBAs are not
validated and notoriously fickle in their reliability. And that's just
how well they predict the extent to which the future
circumstance will again produce the behavior, and that the putative
reinforcer will be available and provided such that it can
be reasonably concluded to maintain the behavior. Now, this is
the thing I think is really important for everybody to

(33:56):
hear when we talk about this. FBAs are an important
process that we engage in for our work. As a
general rule, how people conduct FBAs differs from every single
behavior analyst and how they're in the assessment battery and
the tools they use and how they approach, and that
that's part of why it's so fickle, is like everybody
does it a little bit differently. They use the same tools,

(34:18):
but they do it a little bit differently, and they
do not produce causal relations. They always produce a hypothesis
for what we think the function is, and that we
can never confirm it through a standard descriptive assessment FBA process.
We have to we can only confirm functions through the
functional analysis process and can never confirm it through a

(34:40):
standard FBA process because it's all observational and.

Speaker 1 (34:42):
It's always full of ads.

Speaker 2 (34:44):
Oh, so many ads.

Speaker 1 (34:53):
All right, we're back. Yeah, so we're just ended on
the note that it's very important one too, And I
hope I didn't undermine too terribly by that segue that, Yeah,
like this, you get a hypothesis, you don't get a
confirmed function, and you could potentially attempt to validate that
by applying said intervention and seeing do we see some
shift in the behavior? But it doesn't actually mean you
have identified a clear function. It just means you've identified

(35:15):
a possible reinforcer maybe, or you've otherwise arranged some circumstances
to have mitigated the instance in which this behavior might occur,
which does I think speak to the fact that you
nessy doesn't necessarily always change the motivation for that behavior,
like the reason for it happening, and therefore is still
there's some ethical considerations around is this person getting their
needs met. But let's move on from the descriptive or

(35:38):
observational portion of this and move on to the other
thought of how well behavior could be assessed using just
those questionnaires, those interview tools that are structured. There have
at this point been a handful of various open and
close ended questionnaires and standardized interview tools that have been
developed and standardized to attempt to create valid and reliable

(36:01):
inquiries that may yield useful insight as to what may
be maintaining those challenging behaviors. And again we're like, these
are the people who are around this when it occurs,
They've seen it all the time. Maybe if we just
ask the right types of questions we can get enough information.
And the thinking is that if we administer the same
questionnaire to a handful of people, then the people who

(36:23):
are closest to this individual, and then use their answers,
we can actually make some determines about likely functions of
that behavior because they can we sort of set it
up so they can tell us what are the circumstances
when that behavior occurs, what happens after the behavior occurs,
and it's not those aren't the exact questions, but they
are set up in an attempt to create an opportunity

(36:43):
to assess possible behavior functions just by interviewing the people
who around when that behavior has happened right exactly.

Speaker 2 (36:51):
And while these have demonstrated some utility and clinical settings
for many people in many ways, they also come with
some caveats that they're unreliable. Questionnaires are often unreliable because
different raiders will give different answers to the same questions,
and sometimes the same raider will give different answers if
the questionnaires administered at different times, and notoriously, people are
terrible self reporters, So you know, there's there's also that

(37:14):
piece too, where people are reporting on a phenomenon that
they're very close to they may give different answers depending
on how the question is worded. I mean, it's really
kind of fascinating. I think it was like even one
of the questionnaires that's often used. I don't know if
you talk about this with the functional analysis screening tool.

Speaker 1 (37:29):
I'm not. I did not bring up any specific ones.

Speaker 2 (37:31):
No, okay, I bring I want to bring this up
because I think this is this is an important piece.
Like there's some studies that show that the functional analysis
screening tool, the FAST, which is often used as a
like a screening tool type of assessment. I think they
found that related to functional analyzes that were conducted after
a FAST was completed. I think the FAST is the

(37:52):
most accurate of those types of tools, and it was
like fifty two percent accurate.

Speaker 1 (37:56):
I remember about being fifty yeah.

Speaker 2 (37:57):
Yeah, it's like about fifty percent of the time it
predicts the function. So like that's a questionnaire that is
like it's a coin flip, right, yeah, yeah, at that
point exactly. Yeah, it's like this might tell us something maybe,
but it's hard to tell. Yeah, And to that point,
like just as you were saying, some studies have directly
compared some of the questionnaires against an actual experimental functional

(38:18):
analysis and generally found only moderate correspondence between the experimental
like actual confirmational analysis and the questionnaire, meaning that those
questionnaires will tell you one thing, and when you do
the actual experiment it tells you something different a good
chunk of the time, right, And of course I always
had this thought when they were talking about this, and

(38:38):
I'm like, okay, well, it could be the case that
the questionnaire is the one that's actually yielding the correct
and accurate analysis of the behavior function, and that it's
the FA that is incorrect.

Speaker 1 (38:52):
Yeah, now it does. I mean it could be the case.
It does seem a little bit likely because there is
substantially more research backing the functional analysis and way more
of like validated outcomes that exist there. So it does
make sense that the FA, the standard FA, is the
tool against which other assessments are compared, such as a questionnaire,

(39:13):
to determine are these things yielding behavior function? But just
wanted to throw it out there that it's like when
compared to an FA, that you'll discriver results. And I'm like, well,
that but doesn't tell you. That doesn't technically tell you
which one is right. But given the amount of research
done validating the use of a standard functional experimental functional analysis,
that that is the one that it makes sense that

(39:33):
people will lean on that as the comparison sort of
match to sample, if you will.

Speaker 2 (39:37):
Sure, because I think it's important to understand like maybe
like the process by which we get those data, right, Sure, yeah, exactly,
you know, with like these questionnaires and like with the observations,
like we're seeing things play out, but we're not doing
any sort of like experimental analysis or like manipulating any variables,
like the functional analysis is the one where we do
start manipulating or changing some variable in the in the

(39:58):
environment to see it effect on that response. And I
think that those that is so that's such a critically
different process that I think that when we when we
compare them, it is a little bit like, you know,
the goal is the same to figure out the function
of the behavior, but the way that we get that
information it's just a little bit more robust to the
functional analysis stage.

Speaker 1 (40:19):
Right now. I fully went into this expecting we get
through all that we have left to say about functional analyses,
but there is enough left that I think we actually
do have content for a whole entire other episode because
we still have to get into like what do we
do with the results of a functional analysis? And then
I teed this up last time and we didn't get

(40:39):
to it, but sort of getting into the other innovations
that have happened around functional analyzes, which is specifically synthesized analyses,
and getting into some of the controversies therein and getting
into some of our own sort of hot takes on
some of these issues as well. So rather than drag
this out into an hour and forty five minute long episode,

(41:00):
let's go ahead and stop here. We'll pause here for
this episode and we will go into a part four
to get into those sections as well, which bully for
everyone here. You know, we got so much content out
of this topic. I wasn't sure we'd be like it
this much.

Speaker 2 (41:15):
But who to thunk it?

Speaker 1 (41:17):
Who to thunk it? But okay, perfect, Well, let's go
ahead and pause there. And as we're sort of heading
toward the end of this, for those of you who
have never listened to us before and this is your
first time, then way to pick up a spot right
in the middle. But hopefully you're getting a lot out
of this. Still, at the end of our each of
our discussions, we recommend some things. We are people who
go about the world consuming all kinds of entertainment and

(41:40):
books and movies and music and experiences, and we just
like to share those with people, and so we recommend
them at the end of our discussions, even though they're
not about the topic we were discussing most of the time.
And so we'll get to that. And we also, of
course have the credits. I even have like it's not
technically a listener mail. But it's kind of like a
listener mail i'd like to read, so get into that
as well. But all right, before we go ahead and

(42:03):
make that transition, is there anything that I forgot you'd
like to add before and we can get to credits
and whatnot.

Speaker 2 (42:09):
No, I think we're ready to roll into it.

Speaker 1 (42:12):
All right, Well, we're close enough that I think let's
go ahead and do one more quick ad break and
then we'll come back for the fun wrap up here
where we'll get into our recommendations and all that sounds good. Okay,
we're back. I apologize for the lack of cleverness around

(42:34):
my ad segway, but thank you all for being here.
Thank you for joining us on this journey we continue
to find ourselves on with respect to the weird and
wacky world and deep dive of behavior functions and functional
analyzes with behavior analysts. And how many times did I
say the word behavior and function in this discussion. If
we were playing a drinking game, everyone would be dead.

Speaker 2 (42:55):
Yeah, you would be very sick. You would be hospitalized, So.

Speaker 1 (42:58):
Don't do that. I would like say thank you so
much for listening, thank you for joining us. For those
people who would like to support us, there are many
ways you can do that. You can leave us a
rating and a review on whatever platform you listen to.
Just click the highest amount of stars. That's very helpful.
We do appreciate it. And if you have something that
you'd like to tell us your thoughts about this, you
maybe correct us on something, provide some feedback about something
you maybe didn't like that we said. You don't have

(43:20):
to screw up our algorithm. You can just email us
directly info at wwdwwdpodcast dot com. And I do appreciate
constructive criticism, and of course I accept glowing praise as well.
If you'd like to send that are we do That
is a way that you can do it. If you
like to support us. Other ways you can tell a friends,
make them subscribe to our podcast, or just take their
device and subscribe for them. But you can also take

(43:41):
the plunge and join us on Patreon. If you do that,
you'll get some bonus content, ad free episode, some other goodies,
early episodes, that sort of thing, and whateveryone's really here for.
At the end of each of our full lengths, I
will read the names of the supporters or people who
have been trucking with us for so long, helping us
be a podcast, and we really really appreciate your generosity
and your support and just generally being awesome human. So

(44:02):
a giant shout out and thank you too, Mike m
Mike T, Justin, Kim Brad, Stephanie, Brian, Kiara, and Charlie
for all of your continued awesomeness and support and helping
us be a podcast.

Speaker 2 (44:16):
We're beautiful. You're beautiful, and we love you. We're beautiful
and you love us is what I meant to say. No,
you're beautiful, we love you. Thank you so much for
your time and your attention. You guys are the best.

Speaker 1 (44:26):
And I can say we're beautiful and I love you
because I also have a team of people who help
me make this, who helped the show come to life.
So writing in fact checking from Shane and myself, thank
you for recording with me today, Shane.

Speaker 2 (44:37):
Hey, thanks for having me.

Speaker 1 (44:38):
Our social media coordinator is Emma Wilson, and our person
who does the sounds good is Justin. Our audio producer,
extraordinary God Emperor Penguin, who's so wonderful and great at everything.
He's a delight indeed, and I think that's what I
have to say about the credits. Is there anything that
I forgot?

Speaker 2 (44:57):
No, I think that covers it all.

Speaker 1 (44:58):
Right, let's do a quick kind of a listener mail
like thing. We'll do an auditory segue to this. This
is actually it's a review someone left our podcast, but
it is one of the nicest things anyone has ever
said about us, and so I want to share it.
Was quite flattered when I read it. And we're a
little behind because this is about a month ago. But
we actually had to record a bunch of episodes ahead

(45:20):
of time because we had some scheduling conflicts for several
weeks where we wouldn't be able to record. So we've
been releasing episodes while not recording them, and now we're
back on track here.

Speaker 2 (45:29):
Yeah.

Speaker 1 (45:30):
They titled their review informative and fun and said I
took a class at ASU from my graduate program and
special education ABA and where someone recommended this podcast, and
I thought I would give it a try since I
commute a lot and needed something to listen to. Not
only is it a psychology podcast, but also speaks on
daily political issues and horrors that continue to happen in America.
I enjoy the candid conversation, informative content, and most of

(45:52):
all that I am learning while listening. Shane Abraham have
a great dynamic that reminds me a little of John
Oliver and Last Week Tonight, as the topics are sometimes
to hear but important. Nonetheless, their username is one six
for the Bunch Numbers. So anyway, thank you to String
of Numbers for that very kind and very thoughtful comparison.
Being compared to John Oliver is extremely flattering to me. Yes,

(46:15):
So that was very very sweet of you, and I'm glad.
I also appreciate people who find a love of learning
as we do, and that's kind of what we're up
to in this discussion. So hopefully this has been insightful
as well what we've been talking about.

Speaker 2 (46:28):
Yeah, no, absolutely, and thank you for the kind words.
Anytime you all want to share nice words with us,
that'd be super cool. Yeah, we love that.

Speaker 1 (46:34):
We're here for it, do you get Yeah? All right, Grevy,
Well let's go ahead and transition then to the fun
part here of our recommendations and yeah, cue the music.

Speaker 2 (46:53):
Recommendations. I am going to recommend a place that is
it's maybe a little bit difficult to get to if
you're in the States, but you can at the very
least still donate money to this place, which I think
is cool. And so that is a place called the
Bonarong Wildlife Sanctuary in Hobot, Tasmania. So I bring this

(47:16):
place up because I think this is such a great
example of how sanctuary should be set up and I
think it's really cool. So this place was a fairly
small plot of land in Tasmania, which is the most
southern state of Australia. It's the Island. And for those
of you who don't know, the wildlife in Australia is

(47:37):
very ridiculous. I was just there for a work trip
and I got to experience it. But I say ridiculous
because it's just so unique compared to the rest of
the world. It's just so so different. We found this
wildlife sanctuary kind of like haphazardly. It was last day
we were there and we're like, let's go check this out,
let's see what let's see what this is all about.
And we get there and it was perfect. It was

(47:57):
exactly like I said, what a wildlife sanctuary should be.
Lots of local animals, they have twenty four to seven
care for the animals in terms of like not only
care but for rescues, so like they have a hotline
where people can go pick up animals for rescues, so
it is a rescue as well. They have a hospital
on site that they can do surgery for the animals
and make sure they're cared for, like especially when they

(48:18):
come in with like burns or broken limbs and all that.
And they do educational tours, so you buy a ticket
and they do tours for free. The folks that work
there take you to every animal habitat. They explain the
animals and what they do and what they are all about,
and it's just super fascinating. So to give you an
idea of what they have at this sanctuary, they have

(48:39):
Tasmanian devils. They have a huge area for forester or kangaroos,
which are quite tall in person. You don't realize how
tall kangaroo is until it stands up next to you
and tries to eat food out of your hands. But lovely,
lovely creatures. There were bare nosed wombats, which, by the way,
I fell in love with one named Pebbles, and I
wanted to bring her home to her and spend time

(49:00):
with her. She was very interesting. They short beaked a kidna,
which if you've never seen the kidna, it's like if
an ant eater were a hedgehog.

Speaker 1 (49:08):
It looks what I was actually gonna say, Like I
pulled up the website and they got it and I
was like, oh, kidd does and yeah, they that is
a good way to describe it as an ant eater hedgehog.
That's exactly right.

Speaker 2 (49:18):
Yeah. And the one they had only had three legs,
so like it would like kind of like drop its
butt when it was walking, and I loved it. So
they have batongs, they have gliders, they have qualls, and
they have like a bird place too. So they've got cockatoos, which,
by the way, are just wild there, just like cockatoos
are just flying all over the place. Sure there are
frog mouths, which are really cool birds of prey. There

(49:40):
was a cape Baron geese walking around that had a
condition called angel wing, which meant that couldn't fly very well.
Her name was Lucy, and she liked to greet all
the people that would come in and she was very nice.
But my favorite is that they had these like parrots
called galas. Yeah, and I'm not saying that with an
Australian accent. It was ga la h yeaheah. I'm familiar

(50:02):
and they were very cool too, So yeah, Glas. Anyway,
it's if you get a chance, please donate to them.
They're doing really great work. They're doing great conservation work
and the food was all vegan, so they make sure
that their their snack bar aligns with their goal and
mission of saving animals and conserving the environment.

Speaker 1 (50:22):
Very nice, that's amazing.

Speaker 2 (50:23):
Yeah, it was great.

Speaker 1 (50:24):
Yeah, I realized that you're the galaw clarification was not
for me, that's for audience, But I was and not
a super nerd on this, but that is one of
the It's a really good card in the Wingspan board game.

Speaker 2 (50:37):
So yeah, yeah.

Speaker 1 (50:40):
But but yeah, that's that's cool. Awesome. Well, I now
I feel like I should have gone first because yours
was so much more important than mine.

Speaker 2 (50:46):
But there's pretty important too.

Speaker 1 (50:49):
Ban Wong Wildlife Sanctuary. I'm recommending a movie that recently
came out. It is the new Spider Man movie Brand
New Day. Yes, oh my god, I love this movie.
They have just been absolutely nailing it with the Spider
Man movies in the MCU and regardless of whether or
not it's MCU. And this one was just so extremely

(51:09):
fun and well done and I really really enjoyed it.
Great cast, some big reveals I don't want to spoil
just yet, but action was fantastic. They did they really
actually did some more unique ways of showing like Spider
Man swinging through the city, and like, after all this time,
they're still finding innovative ways to depict what that can

(51:31):
and can look like and how cool it is, Like
it's really neat to see them just continue to push
the envelope in terms of visual effects, in terms of
like action, and just such a well done movie, so
really really enjoyable if you're at all interested in sort
of superhero action movies. This one was super well done.

Speaker 2 (51:51):
I think there's something really special about Tom Holland because
I think he does Peter Parker and Spider Man well.
Like Yes, that's always the challenge of like playing a
character that has like two stinct personalities, like Batman is
like that, like you for somebody to be a good Batman,
you also have to be a good Bruce Wayne right,
and Tom Holland just does both so well. Like Andrew
Garfield was a really good Spider Man. Yeah, but he

(52:11):
was too cool. He was Peter Parker. He was very cool,
and that was not you know, the case. And Toby
Maguire is also another example of like a good Peter
Parker and a good Spider Man, but he always looked
a little bit older. I think that Tom Holland does
a good job of like keeping the youthfulness of Peter
Parker and being a good Spider Man and like the
Hubrist that are not the Hubrist, but like the gravity
that comes along with that character and the tragedy that

(52:33):
comes along with that character. Yeah. I mean, it's just
it's great. It was such a great movie.

Speaker 1 (52:37):
Yeah. So all right, Bonnarong, Wildlife Sanctuary and Spider Man
Brand New Day probably still in theaters at the time
that this comes out, So go check those things out
if you're in Australia or just want to go see
a movie.

Speaker 2 (52:48):
Yeah.

Speaker 1 (52:48):
Otherwise, we have one more part I think to come
back to discuss in terms of functional analysies, So catch
us next week. I think that'll wrap it up with
what with the amount of content we have left, I
do look forward to sharing with everyone, but I think
that's what we have to say. Anything that you like
to add that I forgot before we wrap up.

Speaker 2 (53:05):
Dope, not today.

Speaker 1 (53:06):
All right, this is Abraham and it's Sheene. We're out.

Speaker 2 (53:09):
Siya.

Speaker 1 (53:10):
You've been listening to Why We Do What We Do.

Speaker 2 (53:13):
You can learn more about this and other episodes by
going to WWD WWD.

Speaker 1 (53:18):
Podcast dot com.

Speaker 2 (53:20):
Thanks for listening, and we hope you have an awesome
day

All Episodes

Episode Transcript

Popular Podcasts

Stuff You Should Know

Dateline NBC

The Breakfast Club

.css-15opob5{left:0;position:absolute;top:0.8rem;} All Episodes

.css-14f5ked{margin:0;word-break:break-word;display:-webkit-box;-webkit-box-orient:vertical;box-orient:vertical;-webkit-line-clamp:2;overflow:hidden;}Functional Analysis pt. 3

Episode Transcript

Popular Podcasts

.css-r6mb8g{margin:0;word-break:break-word;display:-webkit-box;-webkit-box-orient:vertical;box-orient:vertical;-webkit-line-clamp:1;overflow:hidden;}Stuff You Should Know

Dateline NBC

The Breakfast Club

All Episodes

Functional Analysis pt. 3

Stuff You Should Know