Le c t u re ( H 1.2 +
3:
psychological testis a systematic
A
procedure for comparing the t wo or more ppl.
behavior of
·behavioral sample (all kinds of behaviorl
collected in systematic way Istandardized conditions)
·
purpose is to delectdifferences btw ppl+cumparing behavior
·
psychometrics is the science concerned with evaluating attributes ofpsych. Tests:
Tupe ofinformation
·
·Reliability
validity
·
Challengesto m e a s u re m e n t :
final resultm ay notbe entirely real
someppare un aware ofthe meaning or implication oftheir results
↳can give distorted results
, Learning goals:Studentcan turn reactions to items into item scores andtest
scores.
n a t u re ofwa r ;a bility:
2 KINDS OF
VARIABILITY:
.
interindividual differences differences between people
-
·
intraindividual differences:differences emerging in one person over timesunder diff. Circumstances
va r i a b i l i t y and distribution of s c o re s
variantedistance from the mean
size ofvariance depends on:
degree to which scores in the distribution differ
·
the metric ofthe in the distribution (cmvs. M) scale ofthe variable
·
scores
+
Factor to consider when interpreting variance:
number is always greater than 0 (no such thing as
negative variance!)
·
no simple small/large interpretation possible
·
context/comparison is needed to interprett h e size
·
mainly importantfor its effects or other values (e.g.:correlation, reliability)
variance always assumes we are dealing with a normal
distribution!
When you go up on variable x,
does variable Y also 90 up?
notdependenton the scale oft he
variables
Standardized!
1 bc ofstandardization
,va r ia n c e - c ova r i a n c e ma rix
+
used to display all relationships
between variables
COMPOSITE SCORES
score thati s made up ofother scores
·
·e.9.:my total score on a questionnaire thatmeasures ADHD is the sum ofmy response to the
questions. We KNOW tHAADHD has two components linattention and hyperactivity / impulsivity)
score:{"
= 4;
composite
·
3 1 4
+
=
variance -
covariance Matrix
1.291 0.400 2
+
+
= 0.400 2.491
=
covariance:0.400
variance, natt.:1.291;Hyperact: 400 I ADHD 2.491
=
calculating variance:
4 adding up all the variances ofall the items on a questionnaire,
also add twice the covariance ofthe items
, Interpreting test s c o re s :
transformation
-
converted standard scores:
4 converting scores with one scale to
Other Scores
1. calculate z-scores
4 I
z*
sDnertmearnew=
4 When dealing with normal
distribution!
calculating percentiles:
·Normal distribution:
iuse z-scores to calculate/10okup p-value
Otherwise:
use frequency table formula!
+
(Fx -
5fx)x100
Px
= .
N
0.5 *10
A
continuity
72% ofppi scored a 4
or lower
currection
a
3:
psychological testis a systematic
A
procedure for comparing the t wo or more ppl.
behavior of
·behavioral sample (all kinds of behaviorl
collected in systematic way Istandardized conditions)
·
purpose is to delectdifferences btw ppl+cumparing behavior
·
psychometrics is the science concerned with evaluating attributes ofpsych. Tests:
Tupe ofinformation
·
·Reliability
validity
·
Challengesto m e a s u re m e n t :
final resultm ay notbe entirely real
someppare un aware ofthe meaning or implication oftheir results
↳can give distorted results
, Learning goals:Studentcan turn reactions to items into item scores andtest
scores.
n a t u re ofwa r ;a bility:
2 KINDS OF
VARIABILITY:
.
interindividual differences differences between people
-
·
intraindividual differences:differences emerging in one person over timesunder diff. Circumstances
va r i a b i l i t y and distribution of s c o re s
variantedistance from the mean
size ofvariance depends on:
degree to which scores in the distribution differ
·
the metric ofthe in the distribution (cmvs. M) scale ofthe variable
·
scores
+
Factor to consider when interpreting variance:
number is always greater than 0 (no such thing as
negative variance!)
·
no simple small/large interpretation possible
·
context/comparison is needed to interprett h e size
·
mainly importantfor its effects or other values (e.g.:correlation, reliability)
variance always assumes we are dealing with a normal
distribution!
When you go up on variable x,
does variable Y also 90 up?
notdependenton the scale oft he
variables
Standardized!
1 bc ofstandardization
,va r ia n c e - c ova r i a n c e ma rix
+
used to display all relationships
between variables
COMPOSITE SCORES
score thati s made up ofother scores
·
·e.9.:my total score on a questionnaire thatmeasures ADHD is the sum ofmy response to the
questions. We KNOW tHAADHD has two components linattention and hyperactivity / impulsivity)
score:{"
= 4;
composite
·
3 1 4
+
=
variance -
covariance Matrix
1.291 0.400 2
+
+
= 0.400 2.491
=
covariance:0.400
variance, natt.:1.291;Hyperact: 400 I ADHD 2.491
=
calculating variance:
4 adding up all the variances ofall the items on a questionnaire,
also add twice the covariance ofthe items
, Interpreting test s c o re s :
transformation
-
converted standard scores:
4 converting scores with one scale to
Other Scores
1. calculate z-scores
4 I
z*
sDnertmearnew=
4 When dealing with normal
distribution!
calculating percentiles:
·Normal distribution:
iuse z-scores to calculate/10okup p-value
Otherwise:
use frequency table formula!
+
(Fx -
5fx)x100
Px
= .
N
0.5 *10
A
continuity
72% ofppi scored a 4
or lower
currection
a
