Block 4 Quiz Flashcards by isabelle courtney

Describe the use of a chi-square test

It measures the distribution between TWO categorical variables & compares the proportions of two or more categorical data sets to see if there’s statistical significance

Independent variable can be two or more groups (i.e treatments)
Dependent variable 2 or less groups (true/false)

How well did you know this?

Not at all

Perfectly

It measures the distribution between TWO categorical variables & compares the proportions of two or more categorical data sets to see if there’s statistical significance

Independent variable can be two or more groups (i.e treatments)
Dependent variable 2 or less groups (true/false)

Describe the statistical test

Chi-squared test

How well did you know this?

Not at all

Perfectly

Key words associated with what statistical test?:

Two categorical variables
Two/+ independent variables
Two/- dependent variables
Compares proportions
Determines statistical significance.

Chi Square test

How well did you know this?

Not at all

Perfectly

Measuring statistical significance of the proportion of adults with lung disease in clinics A, B, & C at a certain point in time

What Statistical test will you use?

Chi-Square test

How well did you know this?

Not at all

Perfectly

Measuring statistical significance of the proportion of people with diabetes in four different ethnic groups.

What statistical test will you use?

Chi-square test

How well did you know this?

Not at all

Perfectly

Describe the use of an ANOVA/F-test

analyzes the variance, it determines if there’s a statistically significant difference between two or more groups by comparing the means.
- Independent variable (manipulated by researcher)
- Dependent (mean)

How well did you know this?

Not at all

Perfectly

What is a one-way F-test/ANOVA?

ONE independent variable
ONE dependent variable

How well did you know this?

Not at all

Perfectly

Mean height of women in clinics A, B, &C

What is the stat test, the independent, & the dependent variables

One way ANOVA/F-test
Independent = Women
Dependent = Mean height

How well did you know this?

Not at all

Perfectly

Mean height of women & men in clinics A, B, &C

What is the stat test, the independent, & the dependent variables

Two-way ANOVA/F-test
Independent (Women & Men)
Dependent (Mean height)

How well did you know this?

Not at all

Perfectly

What is a TWO way ANOVA/F-test?

TWO independent variable
ONE dependent variable

How well did you know this?

Not at all

Perfectly

(Pearsons correlation coefficient)

r = +1

Positive correlation
increase in y = increase in X
(vise versa)

How well did you know this?

Not at all

Perfectly

(Pearsons correlation coefficient)

r = -1

Negative correlation
decrease in y = decrease in x
(vise versa)

How well did you know this?

Not at all

Perfectly

(Pearsons correlation coefficient)

r = 0

NO correlation

How well did you know this?

Not at all

Perfectly

(Pearsons correlation coefficient)

r = 0.9

Weak positive correlation

How well did you know this?

Not at all

Perfectly

what’s the correlation?

r= +1
Positive correlation

How well did you know this?

Not at all

Perfectly

what’s the correlation?

r= -1
Negative correlation

How well did you know this?

Not at all

Perfectly

what’s the correlation?

r= 0
NO correlation

How well did you know this?

Not at all

Perfectly

what’s the correlation?

r= 0.9
Weak correlation

How well did you know this?

Not at all

Perfectly

What is the use of a t-test?

It calculates the difference of the means of two samples & is used to make confidence intervals

How well did you know this?

Not at all

Perfectly

What is the use of a one sampled t-test?

calculates whether sample mean is different then the populations mean

p> 0.05 reject null
p< 0.05 can’t reject null

How well did you know this?

Not at all

Perfectly

What is the use of a two sampled t-test?

calculates whether sample mean of two different groups taken from the same population differs

How well did you know this?

Not at all

Perfectly

What is the use of an unpaired t-test?

For independent samples, it calculates whether sample mean of two different groups sampled at the same time is different

Null = group mean is equal

How well did you know this?

Not at all

Perfectly

What is the use of a paired t-test?

For dependent samples, it calculates whether the sample mean of the same group sampled at two different times is different

Null= group mean is equal at both times

How well did you know this?

Not at all

Perfectly

Describe variance

The sum of the squared deviations from the mean divided by the total number of values

How well did you know this?

Not at all

Perfectly

Variables: Describe independent variable

Doesn't depend on other variables so it can be manipulated by the researcher (i.e dosage of a drug given to a treatment group)

Variables: Describe dependent variable

Depends on other variables so it can't be manipulated (outcome)

Doesn't depend on other variables so it can be manipulated by the researcher (i.e dosage of a drug given to a treatment group)

Independent variable

Depends on other variables so it can't be manipulated (outcome)

Dependent variable

Variables: Describe discrete variables

Whole number values (number of ____)

Variables: Describe Continuous variables

Any number (weight, distance, temperature etc)

Variables: Describe categorical variables

Variables with categories

Variables: Describe Nominal Categorial variables

Categories without any order ( smoker vs non smoker or male vs female)

Variables: Describe Ordinal Categorical variables

Ordered categories (min/mod/severe)

Describe regression

It means making an equation to represent the relationship between a dependent (Y/outcome) & independent (X/exposure) variables

Describe logistic regression

Dependent variable is categorical - Simple (1 independent) - Multiple (more than 1 independent)

logistic regression

It means making an equation to represent the relationship between a dependent (Y/outcome) & independent (X/exposure) variables

regression

Ordered categories (min/mod/severe)

Describe Ordinal Categorical variables

Categories without any order ( smoker vs non smoker or male vs female)

Describe Nominal Categorial variables

Variables with categories

Describe categorical variables

Describe linear regression

Dependent variable is continuous - simple (1 independent) - multiple (more than 1 independent)

r= +1

positive correlation

r= -1

Negative correlation

r= 0.9

strong correlation

r= 0

NO correlation

Describe the z-score

It compares populations with different means & standard deviations

z-score for a 95% CI is

1.96

z-score for a 97.5% CI is

2.24

z-score for a 99% CI is

2.58

Describe confidence intervals

A range of standard deviations that we're ____% sure our data will fall into

How does sample size impact CI?

Larger sample sizes mean narrower CI (small p-value aka more significant)

Wider CI indicates what about CI?

Less statistical significance

Describe a type 1 error

H0/Null is rejected when it's actually true

Type 1 error

Describe a type 2 error

H0/Null isn't rejected when it's actually false

Describe a type 2 error

Describe statistical power

The formula is 1-B (type 2 error) it represents the probability of correctly rejecting the H0/Null

How can you increase statistical power

Increase sample size

Describe what bimodal distribution represents

There are two subgroups in the populations

Describe this distribution

Bimodal, there are two subgroups within the population

Describe what Positive skew distribution represents

A right tailed skew mean < median < mode

A positive skew (right) mean < median < mode

Describe a negative skew distribution

Left tailed mean > median > mode

Negative skew (Left) mean > median > mode

Describe what the standard error of the mean means?

It's the deviation of the sample's mean from the populations mean Increasing standard deviation & decreasing sample both increase chance of SEM!

How can you reduce SEM?

Decrease standard deviation & increase sample size

Formula for SEM

What does standard deviation describe?

It describes variability or dispersion of the data in relation to the mean

68% translates to how many standard deviations

1SD

95% translates to how many standard deviations

2 SD

99.7% translates to how many standard deviations

3 SD

Formula for standard error of proportion

𝑺𝑬(𝒑)=√(𝒑(𝟏−𝒑)/𝒏)

Describe what the standard error of proportion tells us

measure of the precision of our estimate

a small standard error indicates what?

a precise estimate

What stat test?: Numerical data + 1 group

One-sample t-test

What stat test?: Numerical data + 2 groups dependent on each other

Paired t-test

What stat test?: Numerical data + 2 groups independent on each other

unpaired t-test

What stat test?: Numerical data + over 2 groups independent on each other

ANOVA 1independent = One-way anova 2 independents = two-way anova

What stat test?: Categorical data + 1 group

z-test

What stat test?: Categorical data + 2 or more groups independent on each other

Chi-Square test

What is the best way to minimize the chances of a Type I error in a diagnostic ?

Increase the specificity of the test

A researcher is looking at the effect of a new medication upon blood pressure. She concludes that the results of her study are that the medication does affect blood pressure. What best describes her conclusion? a) The results of her study are not valid b) She has accepted the null hypothesis c) The power of her study was >95% d) The p-value for the relationship is <0.05 e) She has accepted the alternative hypothesis

e) She has accepted the alternative hypothesis

A researcher is setting up an experiment. She generates two hypotheses: first, that the medication will not lower blood pressure; and second, that the medication will lower blood pressure. What is her null hypothesis?

The medication will not lower blood pressure

What factors increase power?

1) bigger sample size 2) variability of observations 3) larger effects of interest 4) greater significance (narrow CI/small p-value)

Researchers are looking at whether or not silicone implants cause breast cancer. They are 80% confident tat their sample size of 250 will detect a difference in breast cancer rates, if in fact one does exist. What does this 80% confidence represent? a) The power of the test b) The false positive rate c) An alpha error rate of 80% d) The false negative rate e) A beta error of 80%

a) The power of the test

What's the stat test?: categorical data 2 paired groups

McNemars test

What test: Quantitative variable + 2 groups

t-test

What test: Ordinal variable + 2 groups

Mann Whitney test

What test: Categorical variable + 2 or more groups

chi-square

What test: Ordinal variable + paired data

Wilcoxon test

What test: Ordinal data + more than 2 groups

Kruskal Wallis test

What test Quantitative variable + paired data

paired t-test

What test Quantitative variable + over 2 groups

f-test/ANOVA

A researcher looks at the cholesterol level in men as compared to women. She is trying to determine whether or not there is a significant difference in the proportion of men with elevated cholesterol as compared to women. What is the most appropriate test? a) Regression b) T-test c) ANOVA d) Correlation e) Chi-square

e) Chi-square

A physician is studying the effects of drug A and drug B on cognitive performance in Alzheimer patients. He administers a memory test to two groups of subjects (those taking drug A and those taking drug B) and compares their mean scores. Which of the following statistical tests would be most appropriate for this purpose? a) T-test b) Chi-squared test c) Paired T-Test d) ANOVA e) Multiple regression

a) T-test

A researcher measures the cholesterol level of three demographic groups: Caucasian, Asian, and African American. What statistical test is used to compare the mean cholesterol level of these three groups? a) Regression b) Chi-squared test c) T-Test d) ANOVA e) Correlation

d) ANOVA

Strong positive correlation r= 0.8 to 1

Weak positive correlation r = 0.2 to 0.5

Strong negative correlation r = -0.8 to- 1

No correlation r = 0 to 0.2

Moderate negative correlation r =-0.5 to -0.8

Weak negative correlation r= -0.2 to -0.5

A researcher found that as cigarette smoking increased, cancer rates increased. How are these two variables related?

Positive correlation

A study of 100 women finds that as the cholesterol level increases, the blood pressure does not change. How are these variables related? Zero correlation

Zero correlation