Karteikarten
QTM 100
Karten in diesem Stapel (118)
- 1sample mean (x)the mean of a sample of a population, called a point estimate
- 2sampling variationthe variation in estimates from one sample to another
- 3population distributionthe true distribution of a random variable in a population (what we make assumptions about)
- 4data distributionthe distribution of observed values from a sample
- 5sampling distributionthe probability distribution of a statistic; it represents the distribution of the point estimates based on samples of a fixed size from a certain population.
- 6approximately normal sampling distribution of sample meanwhen the population distribution from which you are sampling is normally distributed
- 7Central Limit Theoremas the sample size increases, the sample distribution of the sample mean approaches an approximately normal distribution
- 8sampling distribution of sampling proportiona collection of sample proportions from a number n of samples
- 9approximately normal sampling distribution of sampling proportionwhen n is sufficiently large such that np>10 and n(1-p) > 10
- 10binomial distributionused for dichotomous categorical variables (has one of 2 possible outcomes)
- 11conditions for binomial distribution1. has 2 possible outcomes
2. the number of trials is fixed
3. each trial is independent
4. each trial has the same probability of success - 12X ~ Bin(n,p)x: number of successes for a random variable
n: number of trials
p: probability - 13X ~ N(u, o)for a normally distributed random variable
- 14probability distributionexplains the random variables
- 15discrete random variabletakes on a set of separate values that can be viewed in a table
- 16continuous random variabletakes on values that are in an interval; view on a curve
- 17normal distributiondescribes the distribution of a continuous random variable with a bell-shaped and symmetric distribution
- 18standard normal distributionspecific case of the normal distribution where mean = 0 and standard deviation = 1
- 19cumulative probabilitythe probability of being less than a value
- 2068-95-99.7% Rule68% of observations fall within 1 sd
95% fall within 2 sd
99.7 fall within 3 sd
(for normal distribution) - 21Z-scoreif given a value x and you need to find probability convert x to a z-score using equation; if given a probability and you need to find value x, find the z-score and calculate using opposite equation
- 22percentilepercentage of observations that fall below a given data point (the area in the probability distribution curve less than the value of interest)
- 23quantilevalue of the data that represents a certain percentile
- 24probabilitythe way to quantify uncertainty or randomness; the proportion of times an outcome would occur in a long run of observations
- 25random phenomenaevents with uncertain outcomes
- 26trialeach time the event or experiment is performed (each flip of a coin)
- 27outcomethe possible results of each trial; there can be multiple
- 28Law of Large Number (LLN)as more observations are collected, the proportion of observations with a particular outcome converges to the true probability of that outcome
- 29sample spacethe set of all possible outcomes
- 30eventsubset of the sample space (e.g. rolling a one on the die)
- 31disjoint eventwhen events do not have any common outcomes (can't happen at the same time)
- 32non-disjoint event2 events that can happen at the same time
- 33General Addition RuleP(A or B) = P(A) + P(B) - P(A and B)
- 34union of eventsconsists of outcomes that are either A or B - use general addition rule to find probability of a union
- 35intersection of eventsevents that are both A and B: P(A and B)
- 36complement rulecan be used to find the probability that something does not happen: P(Ac) = 1 - P(A)
- 37marginal probabilitybased on a single variable
- 38joint probabilitybased on two or more variables
- 39conditional probabilitybased on two or more variables but isolated on a level of one of those variables ( a single row or column): P(AlB) = P(A and B) / P(B)
- 40independent eventsknowing the result of one provides no useful information about the result of another
- 41rules for independent events1. multiplication rule holds true
2. conditional probability rule holds true - 42experimental studyresearchers assign subjects to experimental conditions and the response variable or outcome is observed
- 43observational studyresearchers observe response and explanatory variables without assigning a treatment (non-experimental)
- 44biaswhen the results of a sample are not representative of the population
- 45sample/coverage biasthe sample may not actually be random, can result from sampling method
- 46non-response biasoccurs when the subjects refuse to participate, which generates missing data
- 47response biaswhen subjects give inaccurate answers
- 48sampling framethe list of subjects in the population from which the sample is taken
- 49sampling designmethod used to collect data
- 50simple random sampleeach individual equally likely to be selected as samples (most likely to be representative of population of interest, unbiased)
- 51cluster sampletake a random sample of individuals from within a random sample of clusters (naturally occuring groups in a population)
- 52stratified random sampleselect a random sample of individuals from within each strata (groups made up of similar individuals in a population)
- 53confounding variablea third variable that is associated with both the explanatory and response variable
- 54controlcompare treatment of interest to control group
- 55randomizerandomly assign subjects to treatment and control groups
- 56replicateselect a large sample sizse or replicate the entire study
- 57blockaccount for variables known or suspected to affect the response of interest
- 58placebofake treatment often used as control group
- 59placebo effectchange despite no treatment
- 60blindingexperimental units do not know which group they are in
- 61double-blindboth the subjects and experimenter do not know which group they are in
- 62factorscategorical explanatory variables in a n experiment (conditions we can impose on the experimental units)
- 63blockingcreates groups that are similar with respect to the blocking variable , then treatments are assigned within the block
- 64empirical rulewhen a distribution is unimodal, symmetric, and bell-shaped, this rule can be applied (68-95-99.7??)
- 65potential outlierif an observation falls more than 3 sd's away from the mean
- 66z-scorez= value-mean/ sd
- 67five number summaryincludes minimum value, Q1 (1st quartile), mean, Q3, and max
- 68interquartile rangeIQR - Q3 - Q1
- 69response and explanatory variablesthe response variable depends on or is explained by the explanatory varaible
- 70associationwhen there is a relationship between the response and explanatory variables
- 717 rules for making good figures1. check the data
2. explain color, size, symbol
3. label axes
4. include units
5. keep geometry in check
6. include sources
7. consider uadience - 72dot plotsshow exact data values, more useful for smaller data sets
- 73stem and leafshow exact data values, useful for smaller data sets: all numbers except last on the left, the last number on the right
- 74histogramvalues are grouped into equal width intervals, and height of bars indicates frequency, useful for larger data sets
- 75bar graphfrequency distribution of categorical data
- 76time series plotdata collected over time (time on x axis)
- 77unimodalone peak
- 78symmetricmirror image
- 79bell-shapedfollows a bell shaped curve
- 80bimodaltwo peaks
- 81left-skewedleft tail is longer than the right; skew in the direction of the tail
- 82uniform distributionall values seem approximately equally likely
- 83meansub of obeservations divided by the number of observations (subject to chagne with outliers)
- 84medianmiddle value of the ordered data (resistant to outlierS)
- 85approximately symmetric datawhen mean and median are close together
- 86modevalue of data that occurs most frequently
- 87rangedifference between largest and smallest observations
- 88standard deviationa type of average distance from the mean (quantifies the variability observed in the data)
- 89variancethe square of the standard deviation
- 90anecdotal evidenceconstitutes informal observations
- 91statistical evidenceobservation based on a formal study
- 92focus research question2 groups of interest and the response of interest
- 93populationthe set of all subjects of interest
- 94samplethe subset of the population from which you collect data
- 95censuscollecting data on everyone
- 96descriptive statisticssummarize the collected data through numbers like averages and percentages
- 97inferential statisticsdraw conclusions about a population based on data
- 98parameternumerical summary of a population
- 99statisticnumeric summary of a sample
- 100variablecharacteristic observed in a study
- 101categorical variableeach observation belongs to one of a set of categories
- 102quantitative variableobservations take on numeric values
- 103discrete quantitative variabletakes on a finite number of possible values
- 104continuous quantitative variablehas a continuum of infinitely possible values
- 105dichotomous categorical variablehas only 2 possible outcomes
- 106ordinal categorical variableif categories take on a specific ordering or numbering
- 107frequencythe number of observations in each category
- 108proportionthe number of observations in each category divided by the total number of observations
- 109percentageproportion multiplied by a 100
- 110graphs for categorical variablespie charts and bar plots
- 111point estimatea single number which represents our best guess for the parameter (given by data collected from a sample)
- 112interval estimatean interval of numbers within which the true parameter is believed to fail
- 113confidence intervalan interval containing the most plausible values for a parameter (estimates the parameter)
- 114confidence levelthe probability that this method produces an interval that captures the true parameter
- 115margin of errorthe range of values above and below our point estimate
- 116standard errorwhen we estimate the standard deviation of the of the sampling distribution from sampling statistics
- 117Hsub0 (null hypothesis)represents the status quo - opposite o what we predict (conduct a test assuming this is true, then reject it)
- 118HsubA (alternative hypothesis)statement of effect - what we are hoping to show (opposite of Hsub0)
Informationen
- Beschreibung
- Karteikarten am QTM 100, erstellt von Claire Parrish am 18/02/2019.
- Erstellt von
- Claire Parrish
- Aufrufe
- 0
- Bewertung