331 Stats Exam 3.24.15 - TRUE/FALSE
The sample statistic while estimating the CI for two group proportions is the 'proportion'.
. The sample statistic while estimating the CI for two group proportions is the 'DIFFERENCE'.
In a regression analysis if sum of square error = 200 and sum of square regression = 300, then the coefficient of determination is
0.6. SST = SSR + SSE SST = 500 then use r2=SSR/SST = 300/500 = 0.6
According to the intention to treat principle, after randomization, if a subject drops out from the study, the subject is not included in the final data analysis.
According to the intention to treat (ITT) principle, after randomization, if a subject drops out from the study the subject is included in the final data analysis.
A 99% confidence interval for the true population mean will be narrower than a 95% confidence interval.
False A 99% confidence interval for the true population mean will be wider than a 95% confidence interval.
Three important components of a confidence interval are: population parameter, critical value and the measure of variability of the population parameter.
False, CI = sample statistic +/- CRITICAL VALUE x MEASURE OF VARIABILITY OF THE STATISTIC
In regression analysis, if the sum of square regression (r^2) is equal to the total sum of square, then the regression line is a poor fit for the data
False, (equation: r^2 = SSR/SST) if SSR = SST , then r^2 = 1. When r^2=1, it indicates a PERFECT fit for the data.
A t distribution can be used when the data/observations are skewed.
False, A t distribution can be used when the data/observations are NORMALLY DISTRIBUTED.
A t test for 2 independent groups can be used to compare two group means when the groups are not independent.
False, A t test for 2 independent groups can be used to compare two group means when the groups ARE independent.
A t test for 2 independent groups can be used to compare two group means when the variances of the two groups being compared are not equal.
False, A t test for 2 independent groups can be used to compare two group means when the variances of the two groups being compared ARE equal.
A t test for 2 independent groups can be used when the observations are skewed.
False, A t test for 2 independent groups can be used when the observations are normally distributed.
A z distribution can be used to compare more than two group proportions.
False, A z distribution can be used to compare only two group proportions.
ANOVA compares "between group" mean with "within group mean"
False, ANOVA compares "between group VARIANCE" with "within group VARIANCE"
ANOVA is used to compare 2 independent groups.
False, ANOVA is used to compare 3 or more independent groups.
As sample size (n) increases, df also increases and the t distribution becomes more different from the z distribution.
False, As sample size (n) increases, df also increases and the t distribution becomes more SIMILAR from the z distribution.
Both z statistic and a chi-square test can be used when the independent and dependent variables are measured on the numerical scale.
False, Both z statistic and a chi-square test can be used when the independent and dependent variables are measured on the nominal scale.
Both z statistic and a chi-square test require the assumption of normality to be met
False, Both z statistic and a chi-square test require the assumption of NO normality to be met
Confidence interval is a single value that defines the true unknown population parameter with a certain degrees of confidence.
False, Confidence interval is a 2 numerical value that defines the true unknown population parameter with a certain degrees of confidence.
Correlation can be used to predict the value of the y-variable given a specific value of the x-variable.
False, Correlation can not be used to predict the value of the y- variable given a specific value of the x-variable. Regression is used to predict the value of one variable based on knowledge of another variable.
F test always a two-tailed test
False, F test is always a one-tailed test since S1^2 (numerator) is larger of the 2 variances.
In a chi-square test, if the observed frequency = expected frequency, more likely is the existence of a relationship/association between the variables
False, In a chi-square test, if the observed frequency DOES NOT EQUAL expected frequency, more likely is the existence of a relationship/association between the variables
One-sample sign test is the non-parametric equivalent of a t test for 2 independent groups or a two sample t test.
False, Mann-Whitney U Test is the non-parametric equivalent of a t test for 2 independent groups or a two-sample t test.
Pearson correlation coefficient (r) describes a curvilinear relationship between two numerical variables.
False, Pearson correlation coefficient (r) describes a linear relationship between two numerical variables.
Pearson correlation coefficient (r) is not affected by outliers
False, Pearson correlation coefficient (r) is affected by outliers.
F test should be used if the sample size of both the groups is the same
False, T test should be used if the sample size of both the groups is the same. F ratio is used to test if two variances are equal.
The t has a smaller standard deviation than a z distribution.
False, The t has a larger standard deviation than a z distribution.
When the sample size is < 30 and the population standard deviation is unknown, use of z distribution is appropriate.
False, When the sample size is < 30 and the population standard deviation is unknown, use of z distribution is LESS appropriate. But can still be used
While using the t test to compare two group means, the dependent variable should be measured on the nominal scale.
False, While using the t test to compare two group means, the dependent variable should be measured on the numerical scale.
A circular pattern in a scatter plot suggests a strong correlation between the two variables.
False, a circular pattern in a scatter plot suggests a weak correlation between the two variables. Linear patterns indicate strong correlation.
F ratio can be less than or greater than 1.
False, because the numerator is the larger variance.
Correlation Coefficient (r) value will change if the units of the variables are changed.
False, for correlation it is scale independent, meaning weight (gm) and height (ft) is the same even if weight is measured in kg and height in inches
If the test statistic value in ANOVA (F ratio value) is equal to one, then the samples means are most likely different from each other
False, if the test statistic value in ANOVA (F ratio value) is larger than variance within groups, then the sampels means are most likely different from each other.
Descriptive statistics can help determine if a mean in one group is significantly different from the mean of another group.
False, inferential stats can help determine if a mean in one group is significantly different from the mean of another group.
If you change the units of your independent and dependent variable, the regression equation will remain unchanged.
False, linear regression is not scale independent meaning that regression equation predicting weight from height will change if the units of the variables are changed
For a simple linear regression equation Y=a-bX, Y increases with an increase in X
False, may be either positively or negatively linear.
One-way ANOVA can be used even if the measurements are not normally distributed.
False, must use Krustal Wallis (the non parametric equivalence)
r value greater than one suggests a very strong correlation between the two variables
False, r value is always between -1 and +1.
A t distribution is used to answer research questions about proportions
False. A t distribution is used to answer research questions about means.
Correlation analysis can help determine a cause-effect relationship between two variables.
False. Correlation analysis can not help determine a cause-effect relationship between two variables.
In regression, the line that results in the highest sum of squares of the errors is the best line
False. Highest sum of squares of the errors (SSE) will make it more error prone. Line with lowest sums of squares of error will detemine best fit line.
In regression analysis, if F ratio value is less than or equal to 1, most of the variable in the Y variable is accounted for by regression compared to by error.
False. If F ratio is >>>> 1 that means that there is more accounted by Regression compared to Error. F= MSR/MSE
If zero lies in the confidence intervals for the difference in means of two independent groups, the two group means cannot be equal.
False. If zero lies in the confidence intervals for the difference in means of two independent groups, the two group means can be equal.
In the actuarial method, subjects that withdraw from the study are completely ignored from analysis.
False. In the Kaplan-Meier, subjects that withdraw from the study are completely ignored from analysis
Survival analysis is used to analyze only survival or death data.
False. Survival analysis is used to analyze not only survival/death data, but also data in which time until a specific event is of interest
Kaplan-Meier method gives subjects that withdraw credit for being in the study for half of that interval.
False. The actuarial method gives subjects that withdraw credit for being in the study for half of that interval
If variability "between group means" is >>>>> than variability "within groups", then the means of the sample are most likely not different from one another.
False. that would make them more likely to be different from one another.
95% CI for the true difference in proportion of subjects with an outcome of interest is (-0.22, 0.2). Hence, the proportion of subjects with an outcome of interest can be equal for the two groups.
True
A chi-square test can be used to compare proportions, frequencies and determine an association among variables.
True
A linear relationship between the IV and DV variable is required for a simple linear regression analysis
True
A t distribution is used when we are dealing with numerical variables.
True
Between-group variance tells us how much each group mesan differs from the true population mean
True
DV variable has to be normally distributed in a simple linear regression analysis is required.
True
Degrees of freedom (df) affects the shape of a t distribution.
True
Distance of Y from the regression line remains similar as we go along the line is required in a simple linear regression analysis.
True
F ratio is the larger variance divided by the smaller variance.
True
F ratio is used to test if two variances are equal.
True
For regression analysis, the DV is measured on the numerical scale
True
Higher the value of r2 , more useful is the regression equation
True
If sum of square regression is >>>>> sum of square error, most of the variability in Y is accounted for by the regression line
True
If the correlation coefficient is a positive value, then the slope of the regression line must be a positive.
True
Linear regression can be used to determine a linear relationship and predict the value of the dependent variable from the independent variable.
True
Post-hoc tests following ANOVA allow us to determine which specific groups difer from each other
True
Regression analysis defines a line of best fit that defines the relationship between the variables
True
Successive observations of Y are not related, in other words Y values are INDEPENDENT in simple linear regression analysis is required
True
Total deviation in the Y variable is the sum of deviation accounted for by regression and that accounted for by error.
True
When a subject drops out or is lost to follow up or does not develop the event of interest at the end of the study, the subject is known as a censored observation.
True
With one-way ANOVA, we are trying to conclude that at least one group mean is different from another
True
Within-group variance tells us how much individual measurement in each group differ from the mean of that group
True
A co-efficient of determination value=0.87 means that 87% variability in one variable can be predicted by the other variable.
True.
Pearson correlation coefficient (r) cannot be used to describe a linear relationship when either of the two variables X or Y are not normally distributed.
True.
Regression analysis can be used to predict the value of one variable from the other variable
True.
Regression describes a relationship between variables
True.
F ratio value ̴ 1 means that both the variances (s1 2 and s2 2) are most likely very different from each other.
F ratio value ̴ 1 means that both the variances (s1 2 and s2 2 ) are most likely very the same from each other.
In regression analysis, the variable we are predicting is known as the independent variable or the predictor variable.
In regression analysis, the variable we are predicting is known as the dependent variable or the criterion variable. The known variables are the independent variable/predictor variable.
Mann-Whitney U test is the non-parametric equivalent of a one-way ANOVA
Krustal Wallis is the non-parametric equivalent of a one-way ANOVA
