Statistical Inference

P Value

The probability of _______ results at least as ________ as the ________ results, under the assumption that the null hypothesis is true.

Obtaining, extreme, observed

P value < 0.05

1. What do we do with the null hypothesis?

2. Is it statistically significant?

3. Laymans explanation

NULL = “NO DIFFERENCE”

P < 0.05 = “REJECT THE NULL HYPOTHESIS” (EVIDENCE AGAINST IT)

the observed results are statistically significant

There is likely a difference or assocaition in the population

Unlikely due to random chance

“WE REJECT THE HYPOTHESIS THAT THE TWO POPULATIONS ARE THE SAME” NULL OR ALTERNATIVE ARE NEVER ACCEPTED

P value > 0.05

1. What do you do with the null hypothesis?

2. Is it statistically significant?

3. Laymans exxplanation

A P value greater than 0.05.

“FAIL TO REJECT THE NULL HYPOTHESIS OR THAT THE TWO POPULATIONS ARE THE SAME” ( indicates insufficient evidence to reject the null hypothesis),

The observed results are not statistically significant. “COULD BE DUE TO CHANCE”

“WE DO NOT ACCEPT ANYTHING”

“WE ONLY REJECT THE NULL HYPOTHESIS OR FAIL TO REJECT THE NULL HYPOTHESIS”

Null hypothesis

Starting statement: The null hypothesis is a statement that there is no effect, no difference, or no association between variables in a population.

“NO DIFFERENCE BETWEEN THE GROUPS A AND B”

“THE INTERVENTION HAS NO EFFECT ON OUTCOMES”

“THERE IS NO ASSOCIATION BETWEEN A AND B”

Confidence Interval

_________of values that estimates the _________ value from a _______

Range, true population, sample

95% confidence interval

95% confident that the true population value is within this interval

Confidence Interval contains 0 for differences

“Fail to reject null hypothesis”

NOT adopt or accept it

P > 0.05

No effect

Not statistically significant

Standard deviation

Measure of spread: How spread out a set of data is in relation to it's mean

Statistical inference

Statistical process of drawing conclusions about a population based on a population sample

e.g. P Value and Confidence interval

Population vs Sample

Population = entire group you want to study or draw conclusions about

Sample = Subset of the population used to make conclusions about the population

Types of Statistical inference

Hypothesis Testing

Estimation

Hypothesis Testing

Test to see if the observed effect is likely due to chance , or reflects a true difference or association in the population. “Is it statistically significant?”

P Value and confidence interval tell us about the population(people not part of the study)

true

Null hypothesis

No difference between the groupS

CONFIDENCE INTERVAL GIVE AN INDICATION OF THE SIZE OF DIFFERENCE, WHEREAS P VALUE ONLY INDICATES IF THERE IS A DIFFERENCE

TRUE

A WIDE CONFIDENCE INTERVAL

INCREASES THE % OF CONFIDENCE, BUT LESS PRECISE

NARROW=MORE PRECISE=LESS CONFIDENT

AN EXACT VALUE = 0% CI

100% = INFINITY

INTERPRETING A CI

STATE THE MEASURE FOR THE STUDY E.G. RR OR ODDS RATIO

INTERPRET THE STUDY VALUE

SAY:

1. “THE VALUE IN THE POPULATION IS UNKNOWN” BUT!

2. “WE CAN BE _% CONFIDENT THAT THE POPULATION VALUE LIES BETWEEN A AND B”

IF CI FOR RR OR OR 1, WHAT DOES IT MEAN?

RR = 1 MEANS THERE IS NO ASSOCIATION BETWEEN THE EXPOSURE AND OUTCOME

A CI CONTAINING 1 MEANS THE RESULT IS NOT STATISTICALLY SIGNIFICANT. WE CANNOT REJECT THE NULL HYPOTHESIS OF NO ASSOCIATION

WHAT IS A HYPOTHESIS TEST

USE SAMPLE DATA TO TEST A STATED HYPOTHESIS ABOUT THE POPULATION

H0 = (null hypothesis)

HA cross out = (alternative hypothesis)

determine the p value and statistical significance