8.1 Steps Of Hypothesis Testing

Question 1

Language

Answer 1

It will present itself in “tries” and “chances” g

35 questions on test, each with 5 choices, so if they’re guessing (ho) they can get 1/5 choices right -> 0.20

35 is just N

“Change” is /=

“Majority” is = 50% And “majority” = more than 50% or >.50

Question 2

Ha language

Answer 2

“Thinks this rate has declined” -> they then conduct poll themselves (so first poll was p^ but technically !!p now)
!! But their poll is still p in the HYPOTHESIS equation (every sample is since its a hypothesis FOR p!!)

so ^^ the hypothesis p represents population proportion for entire place EVEN though both Ho and Ha are samples

TRIALS -> “29/50 is their trial run, but the trial was intended with two possible options regardless of !!NUMBER of trials (sample giveaway) -> 1/2 (0.50) , so when 29 /50= .58, they are proving that p>.50”

Question 3

Finding p langauge (trials)

Answer 3

When a trial is initially conducted for a TOI (Ho/ null)
Since p = all POSSIBLE successes/ TOTAL -> when you see the # of successes and total through this trial, dont be fooled! This is the total of TRIALS (part of total) and not the total possibilities (how many options there are in total WITHIN teh trial)<- trial collects TOI for sure and sample trials collect TOI but not accounting for the options

TRIALS -> “29/50 is their trial run, but the trial was intended with two possible options regardless of NUMBER of trials -> 50/50, so when 29 /50= .58, they are proving that p>.50”

Question 4

Remember

Answer 4

In both Ho and Ha, the proportion number is always p, never p^ (so in BOTH, its p= .5 and p> .5)

Question 5

SC - z- score (TEST STATISTIC)

Answer 5

Original formula for test statistic =
X-(x-) <- distance from mean
———
SD <- how many SDs can first in that distance

But we dont have every number for this so- >
STAT
Proportion stats
One sample
With summary
- # of obs
# of successes
Hypo for p- > Ha and Ho

!!! Round to

Question 6

Z score (test statistic) means

Answer 6

It indicates that the observed proportion of TOI (p^/ Ha) is (insert z-score # here) STANDARD ERRORS above or below p/ Ho(MEAN). (Or within 2 SEs or not)

SE=
Within 2 SEs means=

Question 7

Ho and po

Answer 7

They are the same, and if the p^ number (not the consistent Ha p number in teh equation) is different from po , it’s not consistent with po

Question 8

Imp

Answer 8

When you figure out teh researcher’s p^, this isn’t what they are claiming. Like if they get .74 and the p is .75, this doesnt mean they are claiming by THESE numbers that p< .75. Its in teh actual language

“Improve successes rate”
The current accepted rate is p = .73
So p > .73

Question 9

Statistical inference def

Answer 9

Basically the act of the measurement of sample to generalize about p (what we’ve been doing but putting a more general umbrella term to it)

Question 10

Hypotheses def

Answer 10

Statements about population parameters

Question 11

Null hypotheses def

Answer 11

ALWYAS gets benefit of doubt and is assumed to be true throughout the hypothesis testing
(Whoops my dyslixec butt thought I read alternative- just remmebr Gish one is assumed to be true)

Question 12

To see if its SIGNIIFCANCE level def

Answer 12

0.05 or 5%

Question 13

When z-score (test statistic) is away from 0

Answer 13

The more the NULL hypotheses is discredited
Translation : 2 SEs from mean gives MOST reliability , which in turns invalidates the mean point since its establishing a new mena point