Conditions and Parameters for Inference

Question 1

Chi-square homogeneity

Answer 1

1. Randomness: random sample
2. Sample is less than 10% of population
3. All expected counts are at least 5
   df = (rows - 1)(columns - 1)

Question 2

Chi-square independence

Answer 2

1. Randomness: random sample
2. Sample is less than 10% of population
3. All expected counts are at least 5
   df = (rows - 1)(columns - 1)

Question 3

Chi-square goodness of fit

Answer 3

1. Randomness: random sample
2. Sample is less than 10% of population
3. All expected counts are at least 5
   df = num groups - 1

Question 4

1- prop z test or interval

Answer 4

1. Randomness: random sample
2. Sampling dist. approx. normal: 10 successes and failures
3. Independence: 10% condition

df: none

Question 5

1-prop t test

Answer 5

1. Randomness: random sample
2. Sampling dist. approx. normal: large enough sample (>30), parent population is normal, or data is unimodal, symmetric, with no obvious outliers
3. Independence: 10% condition

df = n-1

Question 6

2-prop z test

Answer 6

1. Randomness: random sample
2. Sampling dist. approx. normal: 10 successes and failures
3. Independence: 10% condition
4. Independence: samples taken from two different populations

df: none

Question 7

2-prop t test

Answer 7

1. Randomness: random sample
2. Sampling dist. approx. normal: large enough sample (>30), parent population is normal, or data is unimodal, symmetric, with no obvious outliers
3. Independence: 10% condition
4. Independence: samples came from two different populations

df = smaller n - 1

Question 8

Matched pairs t test

Answer 8

1. Randomness: random sample
2. Sampling dist. approx. normal: large enough sample (>30), parent population is normal, or data is unimodal, symmetric, with no obvious outliers
3. Independence: 10% condition

df = n - 1

Question 9

Linear regression t test

Answer 9

1. Random sample
2. 10% condition
3. Histogram of residuals approximately unimodal and symmetric
4. Residual plot appears random and reasonably linear, with even scatter of points

df = n - 2