Hypothesis testing

Question 1

Estimation

Answer 1

• using the sample data we estimate the distribution of a parameter in the population from which the sample was drawn
• two types of estimation:
• point estimation
• interval estimation

Question 2

Point estimation

Answer 2

-estimate a single value for a parameter that will probably be close to the true value

Question 3

Interval estimation

Answer 3

-find an interval (confidence interval) that has a given probability (the confidence coefficient) of including the true value of the parameter within its specified range

Question 4

Hypothesis testing

Answer 4

-we test the null hypothesis that a specified parameter of the population has a specified value by looking at the sample value

Question 5

Hypotheses

Answer 5

-statements that provisionally link two variables

Question 6

Theories

Answer 6

-sets of definite propositions or facts that are more or less verified already

Question 7

Popper’s logic

Answer 7

-to prove something is very hard, to disprove is relatively easy

Question 8

Null hypothesis

Answer 8

• H0

- this is the statement in opposite which scientists want to disprove or reject

Question 9

Alternate hypothesis

Answer 9

• H1
• gets accepted if the null hypothesis gets correctly rejected
• the alternate hypothesis is what the scientists actually believe in and are trying to demonstrate

Question 10

One tailed hypothesis

Answer 10

• refers to the statement that differences between groups occur in one direction only ie A>B
• in this case, the alternative hypothesis will be A in not greater than B

Question 11

Two-tailed hypothesis

Answer 11

• refers to the statement that differences exist between the two groups but the direction of such differences are not specified ie A>B or B>A but A is not equal to B
• in these cases the alternative hypothesis would be that A=B
• significance level of two tailed hypothesis are halved so larger differences are required to reject a two-tailed null hypothesis
• this is considered more rigourous and is preferred

Question 12

Rejecting the null hypothesis

Answer 12

• it is desirable to reject H0 and approve H1
• use significance level p, which is the probability of rejecting the H0 when H0 is indeed true
• the higher the p, the better the fit between data and H0
• if the p value is low then H0 is not valid and we can reject it if p is very low
• very low p value= 0.05