For example, it is a test of your own personality

After the test, think to yourself what your personality is like How other people evaluate you

Don't blindly believe the results of the test Some tests are very inaccurate!

1

(a) Retest reliability; using the same test, at different times on the same group of subjects administered twice, the correlation coefficient of the two test scores that is the retest reliability, also known as the stability coefficient. Calculation method: correlation coefficient. Precautions: (1) the time interval between the two measurements; (2) the nature of the variables measured; (3) the two test environment should be the same.

(2) Replica reliability; the use of two equivalent but different questions to measure the same group of subjects, and then find the correlation coefficient of the scores of the group on the two tests, also known as the equivalence coefficient. Calculation: correlation coefficient. Note: (1) the two tests are required to be equivalent; (2) the time interval between the administration of the two replicas.

(3) Internal consistency reliability; internal consistency refers to the homogeneity or internal correlation between the items that make up the test.

Methods for estimating the internal consistency coefficient:

(1) Split-half method: In accordance with the general procedure for administering a test, divide all the items into two equal halves (e.g., split in half according to the odd-even number of the items), calculate the total score of each subject on both halves of the test, and then find the correlation coefficient of the scores on both halves of the test.

Correction formulas: (1) Spearman-Brown rtt=2rhh/(1+rhh) (2) Ferranangan (3) Luron

2. K-R method for dichotomously scored tests (1) Cooder-Richardson's formula 20 (2) Cooder-Richardson's formula 21

3. Cronbach alpha reliability coefficients

Estimating internal Considerations for Consistency Reliability

1, the estimated split-half reliability coefficients may be different depending on the method of splitting the halves;

2, the internal consistency of the scale: split-half reliability refers to the consistency between the scores measured in the two halves of the scale, and the alpha coefficient refers to the consistency between the items in the scale and the items in the scale.

3. The alpha coefficients for each subscale of the scale should be calculated separately.

(iv) Rater reliability; Rater reliability is the degree of inter-rater agreement.

Calculation: 1, product-difference correlation coefficient: two raters using continuous variable scores;

2, Spearman rank correlation: two raters using rank scores;

3, Kendall's harmony coefficient: three or more raters using rank scores.