Reading...
Play button
Play button
Use LEFT and RIGHT arrow keys to navigate between flashcards;
Use UP and DOWN arrow keys to flip the card;
H to show hint;
A reads text to speech;
47 Cards in this Set
- Front
- Back
|
Imprecision is the random ____ to get the same result.
Bias is a systematic ____ from the truth. When something is unbiased it is said to be ___. The term observation refers to the unit upon which _____ are made. ex: person, location. Variables are the ____ being measured. |
inability
deviation valid measurements characteristics |
|
|
In a simple random sample of size n, all possible combinations of n individuals in the population are equally ____ to comprise the sample.
|
likely
|
|
|
The shape of a distribution can be discussed in terms of its _____, modality and kurtosis.
Symmetry refers to the degree to which the shape reflects a ____ image of itself around its center. Modality refers to the number of ____ on the distribution. Kurtosis refers to the ____ of the mound. A positive skew is skewed to the ____ and a negative skew is skewed ____. |
symmetry
mirror peaks steepness right left |
|
|
Relative Frequency (%) is the frequency count divided by the ____ expressed as a percentage.
Relative frequencies are important because they do not depend on the ____ of the data. To determine the Cumulative frequency (%) add the current ____ frequency to the ____ cumulative frequency. |
total
size relative prior |
|
|
The mean is highly susceptible to the influence of ____ and skews.
The mean from a frequency table is equal to the sum of (proportion)(x) The mean tells you nothing about the ____ of a distribution.. The median is ____ resistant to outliers than mean. When the mean = median, the distribution is _____. When mean > median there is a ____ skew When mean < median there is a ____ skew. |
outliers
spread more symmetrical positive negative |
|
|
The deviations of a data set will always sum to ____.
A standard deviation of 0 signifies that there is no _____. |
0
variability |
|
|
A random variable is a numerical quantity that takes on different values depending on ____.
Probability refers to the proportion of times an event is expected to ____ in the population. Discrete random variables exist as a ____ set of possible outcomes. Continuous random variables address quantities that take on an ____ continuum of possible values. |
chance
occur countable unbroken |
|
|
Probabilities in a sample space must sum to exactly ____.
In probabilities, the mean is referred to as the ____ value. Risk is the probability of ____ an adverse condition in a specified time period. The relative risk is the probability of disease given ____ divided by the probability of disease given _____. |
1
expected developing exposure non-exposure |
|
|
____ % of the area under the Normal curve lies in the region +/- 1 SD
____ % of the area under the Normal curve lies in the region +/- 2 SD ____% of the area under the Normal curve lies in the region +/- 3 SD |
68
95 99.7 |
|
|
To determine probabilities for normal values that do not fall exactly +/- 1,2 or 3 SD from the mean, we must first ____ the values and then use a Standard Normal table to look up the associated ____. The transformed value is called a ____.
The z-score tells you the distance the value falls from the mean in ____ _____ units. Values that are smaller than the mean will have ____ z scores and are said to fall ___ the mean. |
standardize
probabilities z-score standard deviation negative; below |
|
|
The central limit theorem states that the sampling distribution of means tends toward _____ even when the underlying population is not normal.
The standard error of the mean gets smaller and smaller as the sample size gets ____. Means based on large n are more likely to fall close to the ____ value of the population mean. |
Normality
true |
|
|
The null hypothesis is a statement of no ____. The alternative hypothesis ___ the null.
The objective of the test is to seek evidence ____ the null hypothesis as a way of bolstering the alternative hypothesis. |
difference
contradicts against |
|
|
The statistical distance of sample mean from the hypothesized value of the population mean provides the weight of evidence ___ the null hypothesis. This distance is reported in the form of the ___ statistic (z-stat).
|
against
test |
|
|
Large values of the z-stat represent large statistic differences and thus good evidence ____ the null hypothesis.
The P-value answers the question, "if the null were correct results this extreme or more extreme would ____ y% of the time. When P is small, ____ the null hypothesis. When P is less than or equal to alpha, the results are _____ _____. When an event is is not statistically significant, the evidence is not strong enough to ____ the null hypothesis, thus you accept it. |
against
occur reject statistically significant reject |
|
|
To test the hypothesis that a sample comes from a population with a known mean but an unknown standard deviation, you calculate a ___ statistic.
When the t value is very large, the significance level is very close to ___. When the standard deviation is unknown, the ___ distribution. |
t
0 t |
|
|
A type I error is ____ the null hypothesis when it is _____.
A type II error is ____ the null hypothesis when it is ____. |
rejecting; true
accepting; false |
|
|
Power is the statistical term used to describe your ability to ____ the null hypothesis when it is false. It is a ____ that ranges from 0 to 1.
Power depends on how large the true difference is, the ___ ____, the variance of the difference and the ____ level. |
reject
probability sample size significance level |
|
|
Alpha is the probability of a type ___ error. Beta is the probability of a type ___ error.
Confidence = 1- ____ Power = 1- _____ |
I
II alpha beta |
|
|
Point estimate provides a single estimate of the ____.
The confidence level of a confidence interval refers to the _____ rate of the method in capturing the parameter it seeks. |
parameter
success |
|
|
Synonyms for explanatory variable: X, ____ variable, factor, treatment, ____.
Synonyms for response variable: Y, ____ variable, outcome, ____, disease. A scatterplot's form refers to whether it is ____, curved, or random. The direction refers to an upward trend (____ association), downward trend (____ association) or flat trend (no association). The strength of a scatterplot refers to how ____ data points adhere to an imaginary trend line. |
independent; exposure
dependent; response linear; positive; negative closely |
|
|
The strength of the linear relationship between two quantitative variables is calculated with Pearson's _____ ____. When all the points fall directly on a line with an upward slope, r = ___.
The sign of r indicates the ____ of the relationship (positive or negative). The absolute value of r indicates the ____ of the relationship. r is the ____ product of z-score for x and y. When r is greater than or equal to ___ this is a strong association. When r falls between ___ and 0.7, this is a moderately strong association. When r is less than ____, this is a weak association. The correlation coefficient is influenced by ____. |
correlation coefficient
1 direction; strength average; 0.7 0.3 0.3 outliers |
|
|
Sample correlation coefficient r is the estimator of population coefficient ____.
Testing a correlation coefficient for statistical significance: 1.) Null: rho =0 2.) Find the ___ statistic. T-test = r/SEr. *with n-2 degrees of freedom. 3.) Find the ___ value using the T-distribution. Regression is also used to ____ the relationship between two quantitative variables. A regression model can be used to ____ the value y for a given value of x. Regression is influenced by ____. A residual is the ____ y - expected y. The least squares regression line: y=a +bx. The slope predicts a change in ___ per ____. The intercept is the value of y when x= ____ |
rho
test p quantify predict outliers observed y; x |
|
|
When 0 is captured in the confidence level, the data is not ____ _____.
r squared is referred to as the coefficient of _____. r squared indicates the _____ of the variation in the response variable is accounted for by the _____ variable. |
statistically significant
determination proportion; explanatory |
|
|
In a negative correlation, as the values of one of the variables increases the values of the second variable _____. It is like an _____ correlation.
Bivariate correlation can be used to determine if two variables are _____ related to each other. Using SPSS for Bivariate Correlation: Analyze- correlate- bivariate-options-select means and standard deviations. Three stars indicates very highly significant. |
decreases; inverse
linearly |
|
|
In a negative correlation, as the values of one of the variables increases the values of the second variable _____. It is like an _____ correlation.
Bivariate correlation can be used to determine if two variables are _____ related to each other. Using SPSS for Bivariate Correlation: Analyze- correlate- bivariate-options-select means and standard deviations. Three stars indicates very highly significant. |
decreases; inverse
linearly |
|
|
Regression is based on the concept of least ____. It usually requires both dependent and independent variable to be _____. Single regression addresses a single _____ variable where as multiple regression addresses multiple.
|
squares
continuous explanatory |
|
|
SIMPLE LINEAR REGRESSION:
y=a + bx b= r (standard deviation of y/standard deviation of x) a = (mean of y) - b(mean of x) The population regression Model: y= α+βxi +εxi Standard error of the Regression SY|x = standard error of the estimate Confidence interval = point estimate +/- (t)(SEb) |
EQUATIONS
|
|
|
For ANOVA regression the null hypothesis: the regression line ____ fit the population.
Fstat = Regression MS/ Residual MS Degrees of freedom for regression always equals ____. Degrees of freedom for residual equals n-2. r squared = Regression SS/Total SS |
doesn't
1 |
|
|
When both the independent variable and dependent variable are categorical, use the ____ test.
When both the independent and dependent variable are _____ use simple and multiple linear regression. When the independent variable is ____ and the dependent variable is ____ use the t-test or ANOVA. When the independent variable is ____ and the dependent variable is ____ use logistic regression. |
chi square
continuous categorical; continuous continuous; categorical |
|
|
MULTIPLE LINEAR REGRESSION:
y= α+β1x1 + β2x2 + βixi A dummy variable is also known as an _____ variable and refers to the _____ variable (x) being categorical. This variable is coded with 2 levels: female-male; college-no college; etc. The test statistic has n-k-1 degrees of freedom. |
indicator
independent |
|
|
A binary variable is a categorical variable with ____ possible outcomes.
Prevalence proportions are the proportions of individuals in the population _____ by condition at a particular time. Incidence proportions are also called the average ____. It is the proportion of individuals at ____ who go on to develop a condition over a specified period of time. Binomial random variables are based on counting the number of ____ of n independent trials. |
2
affected risk; risk succsses |
|
|
Testing a proportion for statistical significance:
1.) Ho: p = proportion under the null. 2.) Find the Zstat = sample p -null/SEp 3.) Find the p-value An npq < ____ is a small sample size. The z test is only accurate for proportions in ____ samples. When the sample size is small an ____ binomial test is used (Fischer's exact test). |
5; large
exact |
|
|
The _____ _____ method is the primary method for calculating confidence intervals for proportion. It is based on the Wilson score test.
n= n+4 x = x+2 p = x/n CI = p +/- Z1-alpha (SEp) SEp = suare root(pq/n) |
plus four
|
|
|
One way to compare proportions is in the form of a ____. Comparisons may represent ____ differences or incidence (risk) differences.
The risk difference parameter: P1-P2. The risk difference estimate is synonymous to the _____ risk. In ____ samples, the sampling distribution of P1-P2 is normal. Prevalence ratio and risk ratio are the same as _____ risk. A risk ratio of 1 indicates that risks are ____ in the exposed and non-exposed group (there is no ____ between the explanatory and response variable). |
differences; prevalence
relative large relative equal; association |
|
|
Both the relative risk (p1/p2) and the risk difference(p1-p2) can be used to qualify the ____ of an exposure.
The RR quantifies the effect of exposure in ____ terms. The RD quantifies the effect of exposure in ____ terms. ex: If RR is 2, then the risk in exposed group is 2 times that of the non-exposed. If the RD=0.01114, then there is a 1.1% increase in risk in absolute terms. |
effect
relative absolute |
|
|
Confidence intervals and hypothesis tests are effective in addressing the random ____ associated with sampling and randomization but not ____ errors.
Systematic errors come in these forms: confounding, _____ bias and selection bias. Information bias arises from defects in _____. |
error
systematic information measurement |
|
|
Cross tabulation is used to analyze _____ outcomes.There are 3 different sampling methods.
The relative risk under cross tabulation is prevalence 2/ prevalence 1. The odds of an vent is the proportion of ____ divided by failures. |
categorical
|
|
|
RR=
(disease exposed/total exposed) / (disease non-exposed/ (total non-exposed) OR = (case exposed/total case) / (control exposed/total control) |
EQUATIONS
|
|
|
A chi square statistic is used to test the _____ between the row and column variables. This statistic tests the ____ of variables when data are from naturalistic samples and homogeneity of proportions when the data are from ____ and case-control samples.
The null hypothesis would be: There is no _____ between the row and column variables in the source populations. Alternative: The null is false. A chi square random variable on a normal distribution has ___ degree of freedom. Chi square distributions are ____ and become more symmetrical as their df _____. Chi square statistics are calculated from ____ and expected frequencies. |
association
independence cohort association 1 asymmetrical; increases observed |
|
|
Expected frequency =
(row total x column total)/ table total Chi square = sum(O-E)^2/ E df = (r-1)(c-1) *uses the Z distribution Chi square does not provide the type and ____ of association. |
strength
|
|
|
When observed values = expected values, chi square is equal to ____.
When observed values are far from the expected amounts, chi square ____ and evidence against the null ____. Do not use chi square when more than 20% of cells have an expected value of less than ____. The square root of chi square equals the ____ stat. |
0
increases; increases 5 z |
|
|
When the explanatory variable (x) is ordinal and response variable is binary, it is necessary to test for a ____.
1.) Null: there is no ____ trend. 2.) Find test statistic z Case-control sampling provides an efficient way to study ____ outcomes. The relationship between the explanatory variable and response variable is quantified with an ____ ratio. The CI for testing the odds ratio uses the natural _____. The null Hypothesis: OR=1 which means no ____. When expected values exceed 5, ____ statistic or z statistic may be used. |
trend
linear rare odds logarithm association chi square |
|
|
A test used in a screening program, especially for a disease with low incidence, must have good ____ in addition to acceptable sensitivity.
Screening is performed only when the disease is an important cause for mortality or _____; a proven and acceptable _____ exists to detect individuals at an early, modifiable age; and there is safe and effective ____ available to prevent the disease or its ____. Sensitivity= # of present positive / total present Specificity = # absent negative / # total absent Less specific=less false ____ = more false positives= ____sensitivity More specific=____false negatives=less ____=less false positives |
specificity
morbidity test treatment; consequences negative;more more; sensitivity |
|
|
PPV = # present positive/total # of pos
NPV = #absent negative/ total # of neg Prevalence= cases/ total pop at risk |
EQUATIONS
|
|
|
The region of the hypothesis test that divides the rejection and non-rejection regions is called the _____ value.
When a range is used to estimate a population parameter, it is called an _____ estimate. If the p-value is greater than alpha in a two-tailed test the null hypothesis should be ____. In testing a hypothesis about two population means, if the t-distribution is used, both populations are distributed ____. The purpose of the chi square test is to test for equality of ____. |
critical
interval rejected normally proportions |
|
|
Constructing a mathematical model that can be used to predict one variable by another is called ____.
The Y-intercept represents the ____ value of y when x=0. Residuals represent the difference between the actual Y values and ____ Y-values. When the response variable is categorical and takes on two possible values, this is called ____ regression. SSR/SST = coeff of mult determin. The one-way ANOVA is used to test statistical hypotheses concerning ____. |
regression
predicted predicted logistic means |
|
|
ANOVA is an extension of the ___ test for comparing the means of two _____ populations.
incidence = new cases/pop at risk |
t
independent |
