Exam 3 Flashcards
List the 4 aspects of a good sample
- random
- no bias
- independent
- covers whole population
What is the appropriate statistical test if you want to know whether the average score on this final is significantly different from the known US average of 125 points?
Parametric:
type of tailed test:
non-parametric test:
1 population t-test for the mean
two tailed
1 population sign test
You want to know the distribution of Corbicula in the Potomac River follows a famous distribution called the Hakencamp distribtion
type of test:
tailed direction:
Chi square goodness of fit
one tailed
You survey 1000 people in MD and find out that 50% of them prefer red wine to white wine. You want to compare this to the percent of French win drinkers who prefer red win and know that percent is 72%. Is the percent in MD less than that of France.
type of test
tailed direction:
1 population z-test for percent
one tailed (left)
you want to know if increasing water to a plant causes the plant to grow faster. You have 4 ater treatments. You measure plant growth and can assume the data meets the assumption of normality.
Type of test
tailed direction
Regression or ANOVA
one direction
You have 50 sets of twins available to study the difference between the effects of 2 growth hormones on a person’s growth.
parametric test:
non-parametric test:
2 population paired 2-test
2 population sign test
T/F the F-distribution is exactly equal to the t-distribution when n> 100.
False
All chapter 11 chi square tests have one assumption that you check when analyzing the data, what is it?
Expected number greater than 5
T/F if your R^2 value is close to 1, you know you have a positive relationship between your 2 variables.
False
T/F For randomized complete block studies, you want to create blocks of experimental units that are as similar as possible.
True
T/F For the statistical tests discussed in this class, parametric tests always have more statistical power than nonparametric tests.
true
Describe the z-distribution
- Symmetric?
- Varies with df?
- can be used to analyze results of experiments?
- Yes
- No
- Yes
Describe the t-distribution
- Symmetric?
- Varies with df?
- can be used to analyze results of experiments?
- Yes
- Yes
- Yes
Describe the chi^2 distribution
- Symmetric?
- Varies with df?
- can be used to analyze results of experiments?
- no
- yes
- yes
Describe the f-distribution
- Symmetric?
- Varies with df?
- can be used to analyze results of experiments?
- no
- yes
- yes
Compare and contrast the assumptions of normality for regression and correlation.
Regression:
-only assume data normally distributed in vertical direction
Correlation:
-assume data distributed in both vertical and horizontal directions
T/F If a field has areas of high moisture and low moisture, high sunlight and low sunlight, we should set up the study using independent samples.
False
t/f For ANOVA, if the variability among treatment means is less than the average variability within treatment means, you will reject the null hypothesis.
false
paired/unpaired
you have a 100 seedling that are genetically similar.
unpaired
paired/unpaired
you complete a study on how well a drug reduces blood pressure and you think sex of the subject may influence the results but do not want to quantify that effect.
paired
paired/unpaired
you are doing a study in a greenhouse on plant growth and the amount of sunlight varies in the greenhouse and you want to get rid of the effect of this variation.
paired
define randomization and explain how it is different from random sampling.
Randomization is used to assign EUs to a treatment with an element of chance.
random sampling is used to collect a representative sample from a population using a element of chance.
T/F Pretest versus Posttest studies are usually independent samples.
False
Block or CRD
The effectiveness of Prilosec for treating heartburn is tested by measuring gastric acid secretions in a group of patients treated with Prilosec and another group of patients given a placebo.
CRD
Block or CRD
The effectiveness of Prilosec for treating heartburn is tested by measuring gastric acid secretions in each patient before and again after drug treatment.
Block
Block or CRD
the accuracy of verbal responses is tested in an experiment in which subjects report their weights verbally, this weight is recorded, and then the same person is weighed on a physician’s scale and the actual weight is recorded for comparison.
Block
Block or CRD
The effect of a preservative as an ingredient is tested with a sample of cans of regular coke and another sample of cans of diet coke
CRD
T/F the chi-square distribution is skewed to the right.
true
T/F in a contingency table, the sum of the observed frequencies in a given row equals the sum of the expected frequencies for the same row.
True
What chi square test/tests on qualitative data can be used to analyze the results of experiments.
chi square test of homogeneity
chi square test goodness of fit test
how do you calculate DF for a chi square goodness of fit test?
of categories - 1
T/F the value of chi squared can be negative, zero, or positive
False
T/F in the chi-square test of independence, data are classified according to 2 qualitative variables
true
T/F a hypothesis test involving a goodness of fit is always a left-tailed test
False
T/F when using a chi-square distribution in a hypothesis test for goodness of fit tests, the number of degress of freedom is always the number of cells.
False
T/F in the chi-square test of independence, the total number of observations in either both rows or else in both columns is set by the research
false
T/F the chi square distribution is normally distributed if np and nq > 5
false
T/F Both chi square tests of homogeneity and chi square test of association can be used for 2 qualitative variables
true
t/f for nonparametric tests, instead of examining the actual values in the data set, the relative position or rank of those values is compared instead
True
What is one advantage of using a nonparametric test?
doesn’t require normality of data, easy to do and interpret
T/F for the nonparametric tests discussed in this class, all nonparametric tests have less statistical power than parametric tests
true
t/f for ANOVA, you only do a means comparison test if you reject the null hypothesis
True
T/F the ANOVA test assumes sampling from normal populations with equal variances
True
T/F in single-factor ANOVA, if the null hypothesis is rejected then all of the population means are declared to differ from one another.
False
T/F The F-distribution is symmetrical around the mean zero
False
In single factor ANOVA, when the calculated value of the test statistics F* is great than the table value for F, we will _______
Reject the null hypothesis and conclude that the factor being tested does have an effect on variable
ANOVA is used to compare two or more population _____
means
T/F If x and y are highly correlated, then x is said to cause y to occur
false
T/F if two variables are correlated then the observations will have a pattern of covariance
true
how do you calculate df for a regression analysis
n-2
or F Regression and parametric Correlation tests both have assumptions of normality
true
