Understanding Reliability and Validity in Psychological Testing

Okay guys, this is Dosier once again and we are in our third topic psychological testing and test construction one first

semester and we are today going to be talking about what as psychology students we may not be tired of talking

about a topic you've known since first year or several year and as a matter of fact if there is any

court that have the right to talk about it this course is one of And that is a topic we call reliability and validity.

Yes. Yes. It is not a topic you should move away but because it's a a topic you must

expect in e totality in your exam. So quickly we are still requesting for your

questions at the comment box or your responses and your liking and sharing our videos and please subscribe. Please

subscribe. I'm requesting. I don't know if I sh Okay, very important guys. Very

important. Now, what is reliability? What is validity? Please. And the types I will be very fast with that. But

please expect it. Yes, I'm I'm always trying to emphasize that. I'm also very sure that other schools cannot do

without this topic also wherever you are listening from. So reliability is talking about the consistency of a a

measure. Consistency of a measure across time, across item and across judgment. Reliability is talking about the

replicability, the consistency of that instrument across time, across item and me judgments. I've said it. Let

me I repeated it again. Now what are we saying? We're talking about is the instruments replicability, consistent

replicability, reproducibility and repeatability.

Right? When that instrument when your personality a friend has found you to be consistently honest across different

time in the in the season morning, afternoon, night, Monday, Tuesday, Wednesday, uh January, February, March,

December, right? That is across time, right? Across item. across item we are saying you're honest

here in even during exam time you're honest with money you're honest uh with relationship you're honest I'm using for

example right um across judgment this is not an opinion of your friend alone this opinion of other classmates family

members church members or your religious religion your anything members right your anything yeah so

like I point out across time across item and across judgment these three acrosses gives us the three measure plus another

one that is very important for exam that we're going to be talking about the types of reliability across time has to

do with test rate test reliability across item internal reliability across judgment inter liability and the first

one we're going to be talking about right so what is test retest not to take our time when you are tested now and

tested as it was so it is and ever shall be now uh now what end right okay now that is that is God

so that is test retest um test retest reliability refers to the

the consistency of uh a measure to to uh have a consistent result or have a similar and relatively similar result

across time measures across different time measures. When the result measure of an instrument provides a consistent

uh measure across different time measure that is what about test retest in other

words I use this to measure this and it gave me 98 and I use it next tomorrow it gives me 97 I use it next next tomorrow

96 99 this is a reliable via test returns. Right? Then we talk about uh internal consistency.

If my measure want to measure honesty, like I said, we expect that an honest person will be very uh we play

professionally when you talk about money, right? When you talk about relationship, the person will not look

dubious or will not start sounding dubious, right? Across items the person consistently. So for example, your item

that intend to measure depression, we should see worthlessness, hopelessness, not the one that will at a point tell

you I love to have fun. No, not a depressed person. Right? So across the items you know initially in

the first video for this course I discussed um what is an item right but that is by

the way I've talked about that so when we say across item consistency of that measure to still tap into the same thing

they are tapping into the same item now the interator like I said it is used to validate instruments right we check it

across judgement different experts. This is the when different experts submit similar opinion concerning a measure of

an instrument or a measure or an instrument. We say that instrument has a consistent

internal or let's say interator consistency or interator reliability. Now the last one I wanted to point out

which is very important also as important as even more important if you want for exam right [snorts] split half

correlation. So imagine you had 98 in exam right and okay 90 please let me not suffer myself

with mathematics here. So when you have 19 exam and oh internally you have been having good marks and test retest across

exams you have been having good good ones and different lecturers that interrator has also submitted that you

are trying right now what if split half is conducted now split half is when the 100 questions

that you scored 90 is divided into to 5050. Would you score uh 90% in first half and

say 90% in second half? So in that regards you expected at at least in each of the 50s to score up to 40 to 45

right or 42 uh 48 in the other one. This is a split half correlation

checks whether you are the type that your strength will run down before maybe uh should I say asa no asa is not

doing that way. Yeah I'm looking at there are some clubs that want to rush you

yeah doesn't persona will do it till the end I think. Yeah. Yeah. So um so they will rush with all their strength but

before uh second half 30 minutes 20 minutes to go their dragon their strength is off right. This is is the um

split half want to see that in the two halves of the of the uh measure whether your responses is reliable and

consistent. your A status, a student status whether it's consistent across the two halves then we say this is a

reliable measure of your um outcome or achievement or intelligence. So from here we talk about validity. Quick one.

Validity is when the instrument measures what it's supposed to measure. What he purported to measure. That is when we're

talking about validity. And from there we talk about phase validity by face. when I put up a uh a depression essing

no not ing this time beg depression inventory right I'm expected to see I am hopeless I am worthless

all those kind of by face I'm not even trying to go deep but from the just looking at it by face I'm already seeing

it right that is what we expect we we we see in um face validity right content

validity content please always remember the word holistically holistically if I want to for example

the depression I'm talking about you understand that depression has its um psychological symptoms like feeling of

worthlessness feeling of hopelessness um um others right helplessness those

negative self-t talk as provided by the company triad of TBEC right now about the self about the world about the

future right now what of the physiological [snorts] or we call it the vegetative symptoms

sleeplessness u um hypophasia hypopasia hyperomnia hyperomnia

uh um fatigue or energy retardation those retardations. Now have you checked the

behavioral aspects plus lots of le things social withdrawal or suicidal ideations

or societal attempts. So we expected to check both the thinking component the emotional component feeling components

and the behavioral components. If your measure is not capturing all of these areas then we are saying that it's

lacking in content validity. What is not enough for a wise right? Okay. Now

criterion validity. This is when this your instrument when a measure when an instrument is found to

correlate relate associates with any instrument sorry any measure or any variable or any construct or any

criterion or criteria or uh uh construct is supposed to relate with Yes, I think that is very clear that is

criterion validity. We are saying that this is when this instrument is able to relate. When we say correlate, it could

be a negative one, it be a positive one, it could be a zero correlation. Right? So when your depression instrument is

able to show that I am different from anxiety which is another emotional or whatever or bipolar which another

disorder of a effect this depression inventory is different from this one we say that that criterion validity have

what discriminant validity discriminant criterion validity [snorts] then when it's able to

measure Um like we have beg um no depression anxiety uh scale does.

Yes. Here we you will see that this instrument can measure these things concurrently. We call them concurrent

validity. Please I'm using this instrument I'm talking about just as an example. Right? Please don't quote me.

What I'm trying to explain is the types of criterion validity which is the ones you're going to see in exam hall right

you must have to see them I don't know yeah but when we say a criterium validity of an instrument is able to

differentiate from others that is supposed to in the first place differentiate from or be different from

we say discriminating when is able to go alongside with the ones he should go along with we has what

concurrent validity. Then when he's able to tell you you see this person, wait for the person 20 years time is going to

blow. At that point we say the instrument has predictive validity. If you say this

person will do well in this job, you say this person by this instrument measure is good to go as a a human resource

manager or as a pilot or anything. predictive. You able to predict the instrument is able to predict a future

occurrence of a behavior. We say that is predictive validity and that draws the cutting of this topic and this video

reli subscribe to our channel. It is actually not a reliable situation and never could

it be valid that you have refused to subscribe nor comment or even um give us a thumbs up to encourage us in what we

are doing. Please, I'll be grateful to see those things coming from you guys. Thank you. I seen other videos.