Stats Exam 2
An effect size (d) of what is small?
.2
An effect size (d) of what is medium?
.5
An effect size (d) of what is large?
.8
What influences the power of a study you're about to conduct?
1) Predicted effect size (d) 2) Sample size (N) 3) Alpha level (a) 4) Whether it will be one tailed or two tailed esat*
When a= .10, the chance of making a type 1 error is what?
10%
Probability of making a Type I error, same as significance level
Alpha
The significance level, which is the chance of making a Type I error is called what?
Alpha
Probability of making a Type II error
Beta
Right procedures lead to wrong decisions/conclusions. Ex) deciding the null is false when it is really true
Decision Errors
A measure of the difference between population means. i.e. how much our sample mean differs from the null hypothesis population mean after our treatment
Effect size
What tells how big the effect of the treatment is?
Effect size!
Failing to reject the null hypothesis when it is in fact what?
False; type II error
If the probability is _____, then we assume the result probably just happened by chance. We do not reject the null
High
"ability to crawl under a bar" to establish whether someone's an athlete.. if the bar is much too ______ then i'll conclude that lots of people are athletes who arent What type of error?
High Type 1 error- lots of false alarms!
Knowing that your result (your sample mean) is significantly significant says what?
Highly unlikely sample's mean would come from a population of scores in which the null hypothesis was true. ->That is, your sample mean is unlikely to be due to chance. It is likely due to your treatment
What does setting the alpha really really low do?
Increase the probability of another decision error, called a Type II Error
If the bar is too ____, i'll think no one is an athlete. This is a type 2 error
Low
If the probability is _____, ex)less than 5% of the time, then we conclude that the result was probably NOT due to chance That the result is statistically significant and supports our research hypothesis
Low
What is the probability that your study will find a significant result when the research hyp is true?
POWER
o
SD of known population
Decision errors are possible in hypothesis testing because you are making decisions about populations based on information in what?
Samples
The lower the alpha, the ________ the chance of a Type I Error
Smaller
A significant treatment is not necessarily a meaningful treatment. True or false?
True
True or false? We decide to reject the null hypothesis only if a sample's mean is so extreme that there is a probability (say less than 5%) that we could have gotten such an extreme sample if the null hypothesis is true. But a very small probability is not the same as zero probability.
True!
"Fake findings" get published. One error pollutes the scientific knowledge base.
Type 1 error
The other error delays the discovery if important findings and also hampers researchers ability to increase their status *Null results are barely published
Type 2 error
Rejecting the null hypothesis when in fact it is true; getting a statistically significant result when in fact the research hypothesis is not true
Type I Error
Two types of decision errors?
Type I Error and Type II Error
"false alarm"
Type I error
"Missed it"
Type II
Failing to reject the null hypothesis when in fact it is false; failing to get a statistically significant result when in fact the research hypothesis is true
Type II
rejected the null when it was true rejected the null when it was false
Type II error
What tells you how meaningful your results may be?
effect size
As your N goes up, your standard error ______.
goes down
d
how many SDs your sample is from the pop mean
Whens the only time you calculate your effect?
if the effect is significant
Sometimes our sample mean is sufficiently extreme _____________________________. When this happens and we reject the null, we've made a decision error.
just due to chance
A large or small difference in means is more likely to be a significant result?
large!
.05= more _____ = _____risk for type 2 error
lenient lower
The more _______ your alpha, the ___ likely are you are to commit Type 2 error
lenient, higher less
The _____ the a, the smaller the chance of making a _____________________.
lower; Type I error
The decision is based on ________________.
probabilities
Low power low probability, high power high probability of finding a significant result. right?
right!
large effect size, higher power than smaller right?
right!
What result is merely one that is unlikely to have come about by chance?
statistically significant result
The more _____ your alpha, the more ______ you are to commit a Type II error
stringent/lower likely
A result can have a large effect size but not be statistically significant because the sample size is so small. true or false?
true
The effect size speaks to the importance of the statistically significant result. true or false?
true
When N=10, u=100, and o=15 and a= .05, Say Brainade increased IQ 8 points on the WAIS. Should you reject the null? What's your effect size?
z= M- Mm/ Om (Mean minus mean/ standard error om (standard error) square root of 15 ^2 /N (10) square root of 225/10= 4.74 z= 108-100/4.74 1.68- reject the null