BBH 310 Exam 2 || with 100% Error-free Solutions.
Sample Error correct answers any differences between the population & an individual sample
SEM correct answers An estimate of the average amount of error that's expected for a given
sampling procedure
statistically significant correct answers extent to which a difference between two things exists
OR the likely hood that the difference or relationship we see would happen by chance alone
Testing the null hypothesis correct answers looking to see that there is no relationship
cant directly test for the presence of differences
Look for evidence to Reject the null
P- Value correct answers Probability of observing our findings if the null is true (no difference
between relationships)
If p-value is low correct answers we can say that the probability that there are no differences is
low= REJECT null
p-value < 0.05 correct answers reject the null hypothesis- that there's no difference (there is a
diff) between two groups and conclude that there is a SIG relationship
Type 1 Error correct answers when the researcher rejects the null hypothesis when it should have
been accepted
Type 1 Error aka correct answers False positive
Type 2 Error correct answers Accepting null hypothesis when you should have rejected it
Type 2 error aka correct answers False negative
Null (tested) Disprove correct answers assumes that two variables have no meaningful
relationship, only bits of association that tend to appear at random
Alternative correct answers there are important differences in one variable as another variable
changes
we only believe the differences are real if they are larger than expected
P- Value cheat sheet correct answers assume null
low p-value- reject
high p-value- fail to reject
p<.05- sig. different
p>.05 - not sig different
Risk of Type 2 Errors correct answers higher the state power lower the false negative rate
,Risks..
sample is small
variability is high
effect on predictor outcome is weak
alpha is very small
Risk of Type 1 Errors correct answers Risk of getting few false positive is high when...
alpha is high
many variables are tested
many separate tests are done
A drug doesn't work. But a pharmaceutical company keeps running studies different ways until
finally, one study gets a significant benefit of the drug. They hide the results of the non-
significant studies, and only publish the significant results for everyone to see. The result they
publish is a : correct answers False positive finding, indicating that their way of doing research is
not-specific enough to only find results of drugs that really work
If we set our alpha level (cutoff between significant and non-significant) at 0.05 and run 40
studies on a drug that doesn't work, how many false positive findings should we expect? correct
answers 2
Imagine we have a drug that does help lower blood sugar, but it only helps a little bit, and it
doesn't work for everyone. The researchers are scared of type 2 errors. They come up with a list
of things that they should do to reduce the risk of type 2 errors. Which reason does not belong on
the list, because it will actually increase the risk of type 2 error? correct answers Reduce the
significance threshold alpha
Variable Scale Determines correct answers Choice of Test
Correlation test when... correct answers BOTH variables are quantitative
Chi- Square test when... correct answers BOTH variables are categorial
T- Test when... correct answers 1 variable categorial and one variable is quantitative
when there are only 2 categories or options in the categorical variable
Anova test when... correct answers 1 Variable is categorial and one variable is quantitative
When there are 3+ categories
P-value has a stronger association correct answers the closer it gets to 0
p= .0001 correct answers the result is unlikely to have shown up at random, thus the variable
seems to have a meaningful connection
Correlation correct answers both variable are quantitative
, Correlation null hypothesis correct answers Knowing one variable doesn't help you know the
other
scatter plot will be broadly scattered w/o discernible slope
Correlation alternative hypothesis correct answers If you know one variable you can accurately
predict the other
strongly correlated when r approaches 1 or -1 (inverse)
Correlation t-stat correct answers r
how many standard deviations increase you would expect to get in the outcome variable
Correlation example correct answers man who is 1 SD taller than the average man is .5 SD
heavier than the average
r=.5
Correlation Scatter plot that make... correct answers U, M,O & certain S shape= Uncorrelated
and NOT sig. associated even when they have a relationship
T-test correct answers 1 variabel quantitative and the other is categorial
t-test null hypothesis correct answers the average of the quantitative variable is the same no
matter what the categorial variable is
t-test alternative hypothesis correct answers the mean quantitative variable is different depending
on which category you are in for the categorial variable
T-test p-value correct answers calculated by estimating how likely it is that you can get some
difference between the means by luck when sampling twice from the same distribution and true
population mean
T-stat tells us correct answers how many times bigger is the difference between the means than
the standard error of mean
if absolute value of t is greater than 2 (1.96) = SIG
Example of t-stat correct answers t=500 (very sig)
t= -50 (very sig)
t- 1 or .5 (not sig)
absolute value of t must be correct answers at least 2 (1.96) to be significant at 95% confidence
interval
Anova test correct answers 1 variable quantitative and the other is categorial BUT the categorial
variable has more than 2 categories
ANOVA null hypothesis correct answers all the groups across the categorial variable are
expected to have the same average level of the quantitative variable
Sample Error correct answers any differences between the population & an individual sample
SEM correct answers An estimate of the average amount of error that's expected for a given
sampling procedure
statistically significant correct answers extent to which a difference between two things exists
OR the likely hood that the difference or relationship we see would happen by chance alone
Testing the null hypothesis correct answers looking to see that there is no relationship
cant directly test for the presence of differences
Look for evidence to Reject the null
P- Value correct answers Probability of observing our findings if the null is true (no difference
between relationships)
If p-value is low correct answers we can say that the probability that there are no differences is
low= REJECT null
p-value < 0.05 correct answers reject the null hypothesis- that there's no difference (there is a
diff) between two groups and conclude that there is a SIG relationship
Type 1 Error correct answers when the researcher rejects the null hypothesis when it should have
been accepted
Type 1 Error aka correct answers False positive
Type 2 Error correct answers Accepting null hypothesis when you should have rejected it
Type 2 error aka correct answers False negative
Null (tested) Disprove correct answers assumes that two variables have no meaningful
relationship, only bits of association that tend to appear at random
Alternative correct answers there are important differences in one variable as another variable
changes
we only believe the differences are real if they are larger than expected
P- Value cheat sheet correct answers assume null
low p-value- reject
high p-value- fail to reject
p<.05- sig. different
p>.05 - not sig different
Risk of Type 2 Errors correct answers higher the state power lower the false negative rate
,Risks..
sample is small
variability is high
effect on predictor outcome is weak
alpha is very small
Risk of Type 1 Errors correct answers Risk of getting few false positive is high when...
alpha is high
many variables are tested
many separate tests are done
A drug doesn't work. But a pharmaceutical company keeps running studies different ways until
finally, one study gets a significant benefit of the drug. They hide the results of the non-
significant studies, and only publish the significant results for everyone to see. The result they
publish is a : correct answers False positive finding, indicating that their way of doing research is
not-specific enough to only find results of drugs that really work
If we set our alpha level (cutoff between significant and non-significant) at 0.05 and run 40
studies on a drug that doesn't work, how many false positive findings should we expect? correct
answers 2
Imagine we have a drug that does help lower blood sugar, but it only helps a little bit, and it
doesn't work for everyone. The researchers are scared of type 2 errors. They come up with a list
of things that they should do to reduce the risk of type 2 errors. Which reason does not belong on
the list, because it will actually increase the risk of type 2 error? correct answers Reduce the
significance threshold alpha
Variable Scale Determines correct answers Choice of Test
Correlation test when... correct answers BOTH variables are quantitative
Chi- Square test when... correct answers BOTH variables are categorial
T- Test when... correct answers 1 variable categorial and one variable is quantitative
when there are only 2 categories or options in the categorical variable
Anova test when... correct answers 1 Variable is categorial and one variable is quantitative
When there are 3+ categories
P-value has a stronger association correct answers the closer it gets to 0
p= .0001 correct answers the result is unlikely to have shown up at random, thus the variable
seems to have a meaningful connection
Correlation correct answers both variable are quantitative
, Correlation null hypothesis correct answers Knowing one variable doesn't help you know the
other
scatter plot will be broadly scattered w/o discernible slope
Correlation alternative hypothesis correct answers If you know one variable you can accurately
predict the other
strongly correlated when r approaches 1 or -1 (inverse)
Correlation t-stat correct answers r
how many standard deviations increase you would expect to get in the outcome variable
Correlation example correct answers man who is 1 SD taller than the average man is .5 SD
heavier than the average
r=.5
Correlation Scatter plot that make... correct answers U, M,O & certain S shape= Uncorrelated
and NOT sig. associated even when they have a relationship
T-test correct answers 1 variabel quantitative and the other is categorial
t-test null hypothesis correct answers the average of the quantitative variable is the same no
matter what the categorial variable is
t-test alternative hypothesis correct answers the mean quantitative variable is different depending
on which category you are in for the categorial variable
T-test p-value correct answers calculated by estimating how likely it is that you can get some
difference between the means by luck when sampling twice from the same distribution and true
population mean
T-stat tells us correct answers how many times bigger is the difference between the means than
the standard error of mean
if absolute value of t is greater than 2 (1.96) = SIG
Example of t-stat correct answers t=500 (very sig)
t= -50 (very sig)
t- 1 or .5 (not sig)
absolute value of t must be correct answers at least 2 (1.96) to be significant at 95% confidence
interval
Anova test correct answers 1 variable quantitative and the other is categorial BUT the categorial
variable has more than 2 categories
ANOVA null hypothesis correct answers all the groups across the categorial variable are
expected to have the same average level of the quantitative variable
