![Image](https://www.worldsupporter.org/sites/default/files/styles/medium/public/tips_suggestions/wereldbol_summaries_joho_single_boek_1_150x190px.png?itok=rEljlfYL)
For the differences, indicate which of the two differences below (a or b) you measure:
What conditions does a psychological test need, according to Cronbach, not to meet?
What are malingering and demand characteristics examples of?
A multiple choice exam in which you have only passed or failed is an example of:
Which matters are important when measuring a psychological attribute that is not immediately observable ?
All group 8 students are tested for their reading skills with different assignments. For this they receive a long list with all the words they need to read as much as possible in 5 minutes. What kind of test is this?
The group 8 students are tested for their reading skills for a second time. This time they receice a list of difficult words, where the number of correct read out words is calculated. What kind of test is this?
Psychometrics relates to:
What is, according to Furr, psychometrics?
A = 1, B = 2
D. Test must be taken at different times.
A. Participant reactivity
D. This exam is closed-ended because someone can choose from a limited number of choices. It is criterion referenced because a limit has been set (eg 13 errors is sufficient) that someone must meet to pass the exam.
A. An assumption of relationship between observable behavior and the attribute, this would mean that the test is (at least partially) valid. A theoretical link between the task and the attribute to be measured is important to enable validity. Validity is important because in this case it must be determined with which you can measure the attribute. Operational definitions make it possible to measure and understand unclear concepts (such as intelligence, learning and dreams).
A. Speed tests are tests with a time limit and where you are not expected to complete the entire test and where the test is how far you will respond. This assumes a high probability that answered questions are answered correctly.
B. Power tests are tests without a time limit and in which the answers are actually important, in which the given answer is looked at and based on this a score is determined for the number of correct answers.
A. With psychometrics the emphasis is on the attributes of tests and being able to evaluate this.
B. Psychometrics is the study of evaluating the attributes of psychological tests.
Of what characteristic should a category consist?
Combine the measurement scales (1, 2, 3, 4) with the right examples (a, b, c, d):
Determine whether the statements are true or false.
With an interval scale you can apply the following calculations:
Which of these four examples belongs in an interval scale?
What is the difference between an ordinal scale and a nominal scale?
Measuring with a unit of measurement assumes an important assumption, what is this?
With a driving test, there is a ... variable with which one can pass or fail. The blood type of a test subject is based on a…. variable.
D. The three requirements mentioned are the standard requirements that a category must meet.
All three statements are true.
B. With an interval scale you have an arbitrary zero point, therefore you can only add an substract, but not multiply and divide.
A. At degrees celsius there is an arbitrary zero point, therefore an interval scale is suitable for this.
B. Nominal only indicates difference and interval indicates, in addition to a difference, also a sequence.
A. If the size of the unit of measurement is not always the same, you cannot measure with it because you then provide no / little information.
A. dichotomous has 2 options: (0 or 1 / bags or strokes etc.) and a categorical variable has several categories (not necessarily limited to 2 options).
Of what characteristic should a category consist?
D. The three requirements mentioned are the standard requirements that a category must meet.
What is the difference between variability and covariability?
A test (Test A) is performed in which there are several measurement moments. Every participant takes the same test again at every measurement moment (Test A). The measurement results of the different participants at all times are compared with each other. What differences can be looked at?
Calculate the mean of both rows separately:
Calculate the variance from the following standard deviations:
Calculate the standard deviations from the following variances:
Combine the most likely correlations to the good topics:
1. 0,65 | A. The correlation between alcohol level and the ability to walk straight across a line |
2. -0,65 | B. The correlation between educational level and income |
3. 0,00 | C. The correlation between hair colour and gender |
For a Sum of Squares of 2000 with N=5, what is the standaard deviation?
Complete the table (a t/m e):
Deviation X | Deviation Y | Cross-product | |
A. | 0 | 20 | |
B. | 10 | 10 | |
C. | 0 | 30 | |
D. | 20 | 5 | |
E. | 5 | 10 |
Which correlation list(s) are correct?
A. | B. | C. | D. |
0,5 | 0,5 | 0,9 | -0,9 |
-0,5 | 0,7 | 0,4 | -0,7 |
0,7 | 0,2 | 1,4 | -0,6 |
0,2 | 0,8 | 0,8 | -1,6 |
Is the table below the correct norm table for the raw test score X (M = 80, SD = 12)?
X | z | T |
68 | -2 | 26 |
74 | -1 | 38 |
80 | 0 | 50 |
86 | 1 | 62 |
92 | 2 | 74 |
B. Co-variability is the degree to which the variability between different score sets match. Variability is the difference within a set of scores. Therefore, the subjects are different and there is no question of 1 the same subject in covariability and variability.
C. Both intra-individual differences (the results of a participant over all measuring moments) can be looked at as well as inter-individual differences (between the participants).
S2 is the variance. So to go from s to s2, this just needs to be squared. a is therefore 1002 = 10,000, b is 12 = 1, and c is 25.22 = 635.04.
s2 is the variance. To calculate the variance using the standard deviation, the standard deviation needs to be squared.
s2 is the variance. To calculate the standard deviation using the variance, only the √s2 needs to be calculated.
A.
Sum of Squares / N = 2000/5 = 400 = s2.
√400 = 20.
To calculate a cross product, the deviations can multiply with each other.
List A and B are good, correlations can be between -1 and 1.
D. Both the z-scores and T-scores are incorrect.
What is the difference between variability and covariability?
B. Co-variability is the degree to which the variability between different score sets match. Variability is the difference within a set of scores. Therefore, the subjects are different and there is no question of 1 the same subject in covariability and variability.
What questions are asked about the dimensionality of a test?
The WISC intelligence test is an example of a:
Combine the terms:
Which statement (s) is / are true?
Which statement (s) is / are not true?
When we look at the "eigenvalues greater than one" rule, how many dimensions are there?
Factor | Total |
1 | 1,500 |
2 | 1,200 |
3 | 0,900 |
4 | 0,750 |
5 | 0,450 |
6 | 0,350 |
Which of these statements are true?
In a PCA solution, by far the most variance is explained by the first two components. A picture is made of the 2-component solution, which looks like this:
Which pair of variables is probably the most correlated?
Which of the following statements about principal components in PCA is incorrect?
C. The third question (not mentioned in this question) is if the test has more than one dimension, what are those dimensions?
B. An intelligence test consists of a general factor measured by different sub-factors. These sub-factors each represent different types of intelligence.
A. A test where questions only reflect math skills is a unidimensional test, because only one dimension is measured. With mathematical and reading skills, there are two dimensions, which means that this is a multidimensional test. The WISC is a multidimensional test, becauses it assesses different psychological attributes.
B. Identifying the number of dimensions or factors is the second step in performing an EFA. The first step is choosing a statistical technique.
B. Considering the eigenvalue greater than one rule, you look at the number of values greater than 1. Each of these values higher than 1 count as a dimension. In this particular question there are 2 values higher than 1, which might conclude that the test items reflect 2 dimensions.
C. Both statements are true. CFA is a test to check your existing idea about the test.
D. V2 and V3
A. When choosing a principal component, true score variance is maximized at the expense of error variance.
What questions are asked about the dimensionality of a test?
C. The third question (not mentioned in this question) is if the test has more than one dimension, what are those dimensions?
From which of the factors below is reliability not derived according to the Classic Test Theory?
How is reliability defined in classical test theory?
What is an important assumption about the error?
What are the consequences of the assumption of question 3? You can select multiple answers.
r2oe = 0.4, s2e = 20. What is the reliability?
s2e = 200, s2o = 600. What is the reliability?
Which of the following statements are not true?
What is true?
Sometimes two tests can be mentioned in parallel. Which assumptions do tests need to meet to be called parallel?
RXX = 0.7, S20 = 200. What is sem?
Which of the four models below has the most restrictions (assumptions)?
What is the reliability of a test with a standard deviation of the observed scores 15 and a standard measurement error of 9?
C. Total scores are not included. Reliability consists of true scores, observed scores and measurement errors.
C. The variance of the true scores divided by that of the observed scores.
A. The error has a random effect and is independent of the variables.
A and B.
A. Rxx = 1 - r2oe. So 1 - 0.4 = 0.6
C.
s2t = - s2o - s2e
s2t = 600 - 200 = 400
RXX = s2t / s2o
RXX = 400/600 = 0.67
B and D are not true: Rxx = 1-r20u. And coe is equal to s2e.
C. An error cannot be higher than the total observed score. If there was no measurement error, there was perfect reliability.
C. I and III
B.
RXX= 0,7, s2o=200.
se = sem
s2e= so √(1 - RXX) = 200 √(1 - 0,7) = 109,54
se = √109,54 = 10,46
sem = 10,46
A. With parallel tests, most assumptions are made. In addition to the standard assumptions from the classical test theory, this model also has three extra assumptions.
B.
so = 15 en sem = 9
sem =se
s2o = 152 = 225
s2e = 92 = 81
RXX = 1- (s2e / s2o)
RXX = 1- (81/225) = 0,64
From which of the factors below is reliability not derived according to the Classic Test Theory?
C. Total scores are not included. Reliability consists of true scores, observed scores and measurement errors.
When do two tests meet the criteria for being a parallel test?
If test subjects have a deviating score due to a previous test, can good reliability be measured from a parallel test?
Which of the reasons why the stability consumption of the test-retest reliability can be problematic are real reasons? (Multiple answers may be correct).
What is the consequence of the incorrect use of the stability assumption?
Internal consistency is a practical alternative to "alternate form" and test-retest reliability. Why is this the case?
When looking at Split half estimates, and the rhh = 0.4. What is the RXX then?
What is the reliability when you have this information: rii = 0.6. k = 5?
Which calculation was used for the previous question?
What is not a recognized method of estimating the reliability of a test?
A test of 30 items is split into two parallel pieces of 15 items each. The correlation between the scores on each piece of 15 items is 0.60. What is the reliability of the entire test according to the "prophecy" formula of Spearman-Brown?
B. 1 and 2 are the standard rules for the parallel test, rule number 3 has to do with test-retest reliability.
A. According to the classical test theory, this would mean that the error scores do correlate, so that there is no longer a real parallel test.
All three answers are correct.
C. If the stability consumption is incorrect, you have a score with the change of the true score and the measurement error, so you cannot determine either of these separately. Your reliability cannot be calculated correctly due to these measurement errors.
C. With an internal consistency reliability you only need 1 test at only 1 test moment to determine the reliability.
C.
Rxx = 2rhh / (1 + rhh).
Rxx = (2x0.4) / (1 +0,4) = 0.8 / 1.4 = 0.57.
C.
Rxx= krii / 1+(k-1) rii.
Rxx= (5x0,6) / (1+ (5-1) x 0,6 = 3 / 3,4 = 0,88
C. Standardized alpha coefficient
D. The correlation between the scores on a test before an intervention and those on the same test after the intervention.
B.
Rxx = 2rhh / (1 + rhh).
Rxx = (2x0.6) / (1 +0,6) = 1,2 / 1.6 = 0.75.
When do two tests meet the criteria for being a parallel test?
B. 1 and 2 are the standard rules for the parallel test, rule number 3 has to do with test-retest reliability.
Mind a different notation in this chapter:
What is the formula of the estimated true score?
Rxx = 0,5, so = 5. What is the sem?
What is the 95% confidence interval for the following information?
sem = 5, Xt = 15, Xe = 5.
Which of the statement(s) is/are true?
Which of these consequences of measurement errors is not true?
What does it mean to have a high item-total correlation?
What is compared with the item discrimination index (D)?
What does an item average of 0.85 say in a binary test?
A. Xest = XMo+ RXX (Xo - XMo)
C.
sem = so√(1-Rxx).
sem= 5 x √(1 - 0,5)
sem = 5 x 0,71 = 3.54
For the 95% confidence interval, we use this formula: X 0 ± (1.96) (sem).
The z score with 2 standard deviations (95%) is 1.96. We stick to that in the formula.
sem is 5 so Xo is the only unknown in the formula.
We then know Xt + Xe = Xo.
So Xo in this case is 15 + 5 = 20.
20 ± (1.96) (5). -> 20 ± 9.8. 95% confidence interval: 10.2 to 29.8.
B en C are correct.
D is the correlation that can be calculated from the covariance; A has some mistakes in the notation.
B is wrong, because consequence 4 implies that it is possible to estimate the true correlation between two constructs. There is a formula (the correction for attenuation) in which the true correlation can be estimated if there was no attenuation.
C. That a test item is consistent with the test.
B. Only well-answered questions are included in this comparison.
A. 0,85 indicates the degree to which participants have answered an item correctly and this equals 85%.
Mind a different notation in this chapter:
What is the formula of the estimated true score?
A. Xest = XMo+ RXX (Xo - XMo)
Face validity is:
If someone wants to know if a test is a good measurement method for a construct:
Factor analysis ensures:
Which function does factor analysis not have?
Motivation in our results is correlated with better results, which should also be theoretically the case. There is:
In theory, IQ and happiness are not correlated, there is no connection in the research. There is:
Link linked items to associated factors.The driving test theory is an example of:
Is it true that validity criterion is used to distinguish groups?
Which form of validity is central to research into personnel selection?
The validity of a new questionnaire for depression (N) is examined by comparing it with an existing questionnaire for depression (D) and an existing questionnaire for work attitude (W). The starting point is that depression and work attitude should hardly be related. The following correlations are found: rND = 0.63, rNW = 0.11 and rDW = 0.08 . This indicates:
Assess whether it is correct that the work sample method of personnel selection is based on the following assumptions.
A. Face validity is all about making it seem so related, not whether or not it is.
B. Participants are generally categorized as being the non-experts and when we look at the opinion about the validity of this group we are talking about face validity: whether it initially seems to be that way.
A. In this case, the content is important, and how it looks and how reliable it is is of less importance.
A. Link linked items to associated factors.
D. Both A and B
B. Convergent proof
A. Discriminant evidence describes the extent to which test scores are uncorrelated with tests of uncorrelated construct.
B. Because the measurement is the extent to which the current test score (knowledge about traffic rules ) is correlated with a relevant variable that can only be measured in the future (being able to drive and apply the traffic rules).
A. Yes, validity criterion can divide groups (met and not met a specific criterion).
B. Only convergent validity of N.
C. Both statements are correct
Face validity is:
A. Face validity is all about making it seem so related, not whether or not it is.
Validity generalization research aims at:
What does MTMMM stand for?
What is not an important characteristic of an MTMMM?
Link the types of correlations to the examples:
Correlation | Example |
1. Heterotrait-heteromethod 2. Heterotrait-mono method 3. Monotrait-heteromethod 4. Monotrait-monomethod | A. Social skills measurement through observation and happiness through questionnaire. B. Social skill through observation and questionnaire. C. Feelings of happiness and social skill via questionnaire. D. Feel happiness through questionnaire |
Which statement (s) is / are true?
Calculate the correlation between motivation and general interest.
True correlation is 0.6, Motivation test is 0.81 reliable, general interest is 0.49.
The formula for the correlation between a continuous and a dichotomous variable (rCD) is:
If the Cohen guideline is assumed and there is a reliability of 0.58 ...
If Hemphill 's guideline is assumed and there is a reliability of 0.29 ...
A test for paranoia, a condition that occurs in 2.5 percent of the population, has a sensitivity (ie, sensitivity) of .80 and a specificity of .95. Based on this test, Joop learns that he is a paranoia sufferer, but suspects that his enemies are behind this. What is the probability that Joop actually suffers from paranoia?
What do we mean by the specificity of a measuring instrument that seeks to determine the presence (positive diagnosis) or absence (negative diagnosis) of a certain disorder?
A. Evaluating the predictive power of a test score over different settings and situations.
MultiTrait-MultiMethod Matrices. (what has uppercase letters together forms MTMMM)
D. with a heterotrait heteromethod , the lowest correlation is expected precisely because correlations have been filtered out using the same method and the like.
All four statements are true.
B. rxoyo = rxtyt √ (rxxryy).
rxoyo = 0.6 x (0.9 x 0.7) = 0.6 x 0.63 = 0.378.
A. r CD = c CD / S D
C. ... then this is a big correlation
C. ... then this is a medium correlation
C. 26 - 50%
B. The chance that someone who does not have the condition will get a negative diagnosis.
Validity generalization research aims at:
A. Evaluating the predictive power of a test score over different settings and situations.
Which type of response bias is likely to occur on a multiple choice (MC) test?
With which bias does a participant agree or disagree too quickly with a statement, without fully understanding its meaning?
If a test is not anonymous, there is a greater chance of:
What kind of bias is likely for someone who finds the test to be long and boring?
If someone benefits from, for example, a diagnosis of a condition, which bias could occur?
Which statements are correct?
Which of the aforementioned statements are true?
To which of the four options does the example belong:
A man goes to the doctor for a test of a disorder. A week later he is told that the test shows that he indeed has the condition. The test is incorrect.
F. There is a chance that people will gamble on a multiple choice test.
A. Acquiensence bias means that people respond yes or no or in one direction to a statement quickly without having to think carefully about its meaning.
C. If people take part in a test where their name must be given, there is a greater chance that they will give socially desirable answers.
E. If someone finds the test boring or the test takes too long, there is a chance that they will no longer seriously read and answer the questions. The person will then answer randomly or give answers that have nothing to do with his or her own opinion / experiences.
D. Exaggeration of problems (Malingering) is a phenomenon in which someone pretends that his or her brain injury or condition is worse what is actually the case, because the person can personally gain from it. Surcharges, benefits, other care or compensation can be seen as examples for such personal gain.
All statements are true.
C. Both are true.
B. a false positive: the result appears to be positive but this result is not justified.
Which type of response bias is likely to occur on a multiple choice (MC) test?
F. There is a chance that people will gamble on a multiple choice test.
What is not possible with construct bias?
Prediction bias is when ...
Which statement is true?
I. Item discrimination index can be used to discover construct bias.
II. Differential item function analysis can also be used to discover construct bias.
III. Factor analysis can be used to evaluate the internal structure of a test, separately for two groups.
Are these statements true or false:
Is a difference in test scores between groups sufficient reason to assume that there is a bias? (possibly multiple answers)
Which statement is true?
I. Two separate groups in a test do not show the same internal structure for their test scores. We can therefore conclude that the test suffers from construct bias.
II. Two separate groups in a test display have the same internal structure for their test scores. We can therefore conclude that the test does not suffer from construct bias.
What is not a method to detect construct bias?
Are these statements true or false?
Are these statements true or false?
In which bias is difference consistency involved?
What type of proof is not relevant for construct validity?
A. Because construct bias means that scores on a test can have different meanings for different groups. They cannot be properly compared with this reason.
A. The relationship between true and observed score differs between two groups.
A. All statements are correct.
Only statement A is true, because the types of bias at B and C have been reversed.
B and C are right, A. is not, this is a too fast and unequivocal conclusion.
A. Only statement I. is true
D. Rank order is a method to discover construct bias but item order is not.
A is true, B. not: the regression formula for the common regression line is based on data from all groups, not just 1 group.
Yes, only the intercept differs between the groups, the direction coefficient is the same.
A. With intercept bias, the difference remains constant while the x rises or falls. This is not the case with the rest.
A. Face validity of the items
What is not possible with construct bias?
A. Because construct bias means that scores on a test can have different meanings for different groups. They cannot be properly compared with this reason.
CFA can be used to evaluate hypotheses about the internal structure of a measurement model. But what are the steps you need to take before you can start?
I. Score negatively coded items in reverse.
II. Make clear which construct you are going to measure + already develop number of items.
III. Find participants.
IV. Already make the test yourself.
If a CFA is performed, what should be established first?
A factor load is:
In the fourth step, executing the CFA with software provides information about:
What does a significant chi-square say about the assumed model?
Which of the factor loads has the least chance of remaining in the test?
How many lower-order factors does the model have?
Which of the " lower-order" factor has / have the bad load on the " higher-order" factor?
Are the statements true or false?
What is a standardized residual in a confirmatory factor analysis?
D. You don't have to take the test yourself, the rest is important to have done before you start performing the CFA.
First (after entering the data) the number of dimensions must be determined.
B. This is the extent to which an item is associated with a factor. This factor loading is the first parameter, the second parameter is the connection between different factors.
B. Fit or suitability of model
A. It may indeed mean that the sample was large (options C and D), but that does not necessarily have to be the case. Moreover, a significant chi-squared means a poor fit, and therefore a poor agreement with the actual results. (B and D are not possible)
C. The lower the factor load, the worse the reflection of the underlying factor, which means that it is most likely to disappear from the test.
B. The second to last column is here the column with " lower- order" factors. This is the column after the highest factors for the items.
A and B, these are the lowest values (- or + does not matter).
A and B are true. C is not true, because this method can be used for both discriminant and convergent validity. D is true; we can evaluate convergent validity by examining a test and one or more criterion variables using CFA.
D. The difference between the fit measure CFI and the standard fit fit NFI.
CFA can be used to evaluate hypotheses about the internal structure of a measurement model. But what are the steps you need to take before you can start?
I. Score negatively coded items in reverse.
II. Make clear which construct you are going to measure + already develop number of items.
III. Find participants.
IV. Already make the test yourself.
D. You don't have to take the test yourself, the rest is important to have done before you start performing the CFA.
Which of the claims are true?
If there are five items per test and one observer, how many facets are there?
There are five items in a test and the test is measured at two moments by two observers. How many facets are there in this study?
Are these statements true or false?
Which of the claims are true?
What is the target variance if there are 4 items, MSt = 5, MSres = 1 and MSi = 2.
There is a generalizability coefficient of 0.790. What is the noise when the signal is 50?
Are the following statements true?
Combine the decision type with the corresponding example and with the corresponding test type:
1. Relative decision | A. Theory driving | I. Norm-referenced test |
2. Absolute decision | B. Auditions | II. Criterion-referenced test |
B and D are true.
A. There is 1 facet: the items. For the rest of the measurement options, there is only 1.
C. There are 5 items (1 facet) at 2 moments (2nd facet) by 2 observers (3rd facet)
A is true, B is not, this happens in the first step of the G theory analysis (the G study).
D. is false, the rest is true. Measurement errors can influence variability.
C.
Target variance if there are 4 items, MSt = 5, MSres = 1 and MSi = 2:
(MSt-MSres) / Ni. Ni = 4, (5-1) / 4 = 4/4 = 1.00.
There is a generalizability coefficient of 0.790. What is the noise when the signal is 50? The generizability coefficient is signal / ( signal + noise ). The formula can be reversed to discover signal . 50 / (50 + x) = 0.790. → times (50 + x) → 50 = 0.790 * (50 + x). → divided by 0.79 → 63.29 = 50 + x. → - 50 → X = 13.29. Noise is 13.29.
Both A and B are true.
1-A-II.
2-B-I.
Relative decisions are about relatively assessed scores and are norm- referenced (e.g., best 20%). Absolute decisions are about a cut-off score that you must have to have achieved something, this is criterion-referenced.
Which of the claims are true?
B and D are true.
Which of these statements are not true?
For which demand is probably an higher traitlevel required when the trait is about 'proficiency in spelling'?
Are the statements true or false?
There is a negative discrimination value of the item. What does this mean for the chance to answer an item correctly when there is an high trait level?
Which discrimination value has the most validity?
The Rasch model looks at:
(possibly several correct answers)
The difference between the Rasch model and the two-parameter logistic model is:
Enter the formula of the Rasch model: Trait level = 2, Item difficulty = 1.5. What is the chance of a correct answer?
What is the estimated trait level if the proportion of correctly answered items from the respondent is 0.7?
Which of the following examples is NOT a named application of IRT?
A test consisting of four items measures a certain skill. The item-characteristic curves of the items are given.
Which item has the largest discrimination parameter?
A test consisting of three items measures a certain skill. The item-characteristic curves of the items are given.
What is the most likely score pattern (order item 1, item 2, item 3, where 1 = right and 0 = wrong) for a person with a skill of θ = 6?
B is not true; IRT is more complex than KTT.
B. Assuming that this has an higher degree of difficulty than the assignment for answer A.
A is true, B is not true; more chance than 50%, C is true.
D. That chance is low because a negative discrimination value means that high trait scores result in a smaller chance to answer the item correctly.
A, because that is the highest positive score (only positive scores confirm validity) negative indicate correct inconsistency between trait level and chance of correct answering, this indicates more about poor validity.
A. This is the out1.1.0come of the model and b + c: which is included in the calculation.
A. The discrimination parameter of the item is the difference between the two formulas, they both take the trait level and difficulty of the item into account.
D.
Trait level is Ө s, Ө s = 2. Item difficulty is β i, β i = 1.5.
P (Xis = 1 | Ө s, β i) = (e ^ ( Ө s - β i) / (1 + e ^ ( Ө s - β i)) P (Xis = 1 | Ө s, β i) = e ^ (2-1.5) / 1 + e ^ (2-1.5) P (Xis = 1 | Ө s, β i) = e ^ (0.5) / 1 + e ^ (0.5 ) P (X 18 = 1 | Ө s, β i) = 1.6487 / 2.6487 = 0.622
C.
Proportion correctly answered item of the respondent = PS. PS = 0.7.
Ө s = LN (PS / 1-PS).
Ө s = LN (0.7 / 0.3) = LN (2.333) = 0.897.
B. this is a function of the G theory not of IRT.
A. Item 1
D. 1.1.0
Which of these statements are not true?
B is not true; IRT is more complex than KTT.
JoHo can really use your help! Check out the various student jobs here that match your studies, improve your competencies, strengthen your CV and contribute to a more tolerant world
There are several ways to navigate the large amount of summaries, study notes en practice exams on JoHo WorldSupporter.
Do you want to share your summaries with JoHo WorldSupporter and its visitors?
Field of study
International Bachelor Psychology IBP Gioia Grosso contributed on 06-11-2023 23:49
Wanting to pass the Psychometrics exam
Add new contribution