Radford University



Graduate Research Methods Homework # 4 (10 pts)

The data below represents the pool of responses obtained by giving the same test to a group of participants with a 3 month delay between presentations.

T1 Subj. 11 21 31 T2 Subj. 12 22 32

Question 1 2 5 6 2 1 3

Question 2 2 5 3 2 1 7

Question 3 1 4 1 1 1 4

Question 4 1 3 5 1 1 2

Question 5 3 1 2 1 2 1

Sum 9 18 17 Sum 7 6 17

Mean 1.8 3.6 3.4 Mean 1.4 1.2 3.4

1. Calculate the test retest reliability.

2. Calculate the split-half reliability for time 1 and time 2 separately. Use the Spearman-Brown Prophesy formula to estimate the reliability for the whole test. Note: ignore the fact that the # of odd and even variables are not equal.

3. Using the standard error of the difference, estimate whether subject 2's score for time 2 significantly differed from time 1. Use the SD from test 1 and the Spearman-Brown corrected split half reliabilities from problem 3.

|Time 1 |Q1 |Q2 |Q3 |Q4 |Q5 |

|Question 1 | |0.577 |0.277 |0.961 |-0.721 |

|Question 2 |0.577 | |0.945 |0.327 |-0.982 |

|Question 3 |0.277 |0.945 | |0 |-0.866 |

|Question 4 |0.961 |0.327 |0 | |-0.5 |

|Question 5 |-0.721 |-0.982 |-0.866 |-0.5 | |

| | | | | | |

|Intercorrelation Average | | | | | |

|Mean rij (Theoretical Alpha) | | | | | |

|Alpha (Adjusted) | | | | | |

|Mean rij if item deleted | | | | | |

|Alpha (adj) if item deleted | | | | | |

|Time 2 |Q1 |Q2 |Q3 |Q4 |Q5 |

|Question 1 | |0.933 |0.866 |0.866 |-0.866 |

|Question 2 |0.933 | |0.988 |0.988 |-0.629 |

|Question 3 |0.866 |0.988 | |1 |-0.5 |

|Question 4 |0.866 |0.988 |1 | |-0.5 |

|Question 5 |-0.866 |-0.629 |-0.5 |-0.5 | |

| | | | | | |

|Intercorrelation Average | | | | | |

|Mean rij (Theoretical Alpha) | | | | | |

|Alpha (Adjusted) | | | | | |

|Mean rij if item deleted | | | | | |

|Alpha (adj) if item deleted | | | | | |

4. Complete the Alpha tables above. What would alpha be if item 5 was reverse scored at time 1 and time 2?

5. Two raters have coded 51 participant’s behavior using a 4 point ordinal rating system. Use Cohen’s Kappa to establish inter-rater reliabilities. (The data appear below).

| |R1 1 |R1 2 |R1 3 |R1 4 |

|R2 1 |n=8 |n=2 |n=1 |n=2 |

|R2 2 |n=3 |n=6 |n=0 |n=1 |

|R2 3 |n=2 |n=1 |n=8 |n=4 |

|R2 4 |n=2 |n=1 |n=1 |n=9 |

The data below, represent student scores on an exam of their knowledge of south park trivia. Individual item scores represent whether the student got the question correct (1)or incorrect (0).

6. Estimate this exam’s reliability using KR21.

7. Estimate this exam’s reliability using KR20.

|Kuder-Richardson 20 | | | | | | |

|Questions |Stan |Kyle |Kenny |Eric |Butters (aka: Not Kenny) |

|Who shot Mr. Burns |1 |0 |1 |0 |0 | |

|Who invented Cheezy Poofs |0 |1 |0 |1 |0 | |

|What is Chef's real name |1 |1 |1 |0 |1 | |

|Why shouldn't you do drugs |1 |0 |0 |1 |0 | |

|Will Pig and Elephant DNA splice? |1 |1 |1 |0 |1 | |

|Student Score (X) |4 |3 |3 |2 |2 | |

................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download