MIdterm 1 Study Guide
MIdterm 1 Study Guide 1350
Popular in Elementary Statistics
Popular in Stats
This 2 page Study Guide was uploaded by Rachael Kroeger on Monday May 18, 2015. The Study Guide belongs to 1350 at Ohio State University taught by Strait in Spring 2015. Since its upload, it has received 107 views. For similar materials see Elementary Statistics in Stats at Ohio State University.
Reviews for MIdterm 1 Study Guide
Report this Material
What is Karma?
Karma is the currency of StudySoup.
You can buy or earn more Karma at anytime and redeem it for class notes, study guides, flashcards, and more!
Date Created: 05/18/15
STATISTICS a collection of topics related to collecting and analyzing data DATA a list of values for different variables associated to different individuals 0 The objects that are described by a set of data data analysis learn about population using data from small sample taken from that population VARIABLE any characteristic of an individual can take different values for different individuals 0 Quantitative variable takes numerical values for which being able to do math makes sense ex height age people on high street 0 Categorical variable places an individual into one of several different categories or groups ex major hometown eye color gender OBSERVATIONAL STUDY observe individuals to measure variables of interest without attempting to in uence a response sit back and watch ex surveycensus EXPERIMENT deliberately imposes some treatment on individuals in order to observe their responses make individuals do something in particular POPULATION entire group of individuals about which we want information SAMPLE part of the population that we actually examine to gather information for purpose of drawing conclusions about whole population PARAMETER a number that describes the population ex avg weight of population of dogs in cbus STATISTIC number that describes a sample ex avg weight of a sample of 500 dogs in cbus estimate parameter by a sample statistic Population Inference Sample parameter Sampling statistic CONVIENCE SAMPLE made up of individuals that are easiest to reach giving a survey to friends BIASED VOLUNTARY RESPONSE SAMPLE people who choose to be in the sample themselves by responding to a general appeal online poll BIASED RANDOM SAMPLE best sample reduces bias 1 SIMPLE RANDOM SAMPLE SRSz each individual has an equal chance of being chosen draws names from hat start wpop list random method used to select n individuals from list 2 STRATIFIED RANDOM SAMPLE individuals in the same group have an equal chance of being chosen divide individuals of pop into groups based on some characteristic SRS win each group combine all into big group Stat 1350 Fr SR Students Soph SR El Jr I SRS combine for NONRESPONSE ERROR sample failure to obtain data from an I SF I 5R5 individual selected for the BIAS consistent repeated deviation of the sample statistic from the population parameter true value in the same direction when taking many samples reduce by random sample VARIABILITY how spread out the values of the sample statistic are when we take many samples depends on sampling design and sample size n reduce by increasing sample size MARGIN OF ERROR MOEz tells how close the estimate sample statistic is to the truth population parameter MOE is way to measure variability sample occurs if subject cannot be contacted or refuses to participate EXPLANATORY independent VARIABLE explains or causes changes in the response variable RESPONSE dependent VARIABLE measures outcome or result of study TREATMENT any speci c experimental condition applied to the subjects if experiment has several explanatory variables a treatment is any combination of speci c values of MOE reported with con dence level these 95 LURKING VARIABLE has important 1 effect on relationship among the variables in a study but is not included as explanatory variable CONTROL control effects of lurking size variables on the response by ensuring Gives decimal convert to percent all subjects are affected similarly by these variables then Simply compare We are 95 con dent that between two or more treatments and of all individuals are what RANDOMIZATION39use 39mpersonal were tested lt chance to aSSIgn subjects to treatments AS sample size increases MOE so treatment groups are similar on MOE W n is the sample decreases average at the begging of the study As pop increases MOE doesn t helps reduce b39aS change USE MANY SUBjECTS helps reduce variability PLACEBO dummy treatment with no active ingredients response placebo effect SIMPLEST DESIGN EXPERIMENT Subjects Treatment Results Problems not changing anything not comparing this treatment to anything elselurking variables placebo effect bias COMPLETELY RANDOM DESIGN all subjects are assigned randomly to different treatment groups without accounting for any other variable before hand RANDOM BLOCK DESIGN group of experimental subjects that are known before the experiment to be similar in some way that is expected to affect the response to treatments 0 Blocks can be larger groups of many subjects and can be much smaller units Random assignment of subjects to treatments is carried out separately within SAMPLING ERROR error caused by the act of taking sample Cause sample results to be different from if census was taken reduce by xchange sampling method 0 UNDERCOVERAGE occurs 39 when some groups in pop are left out of sample selection process 0 Sampling frame list of individuals that a sample is drawn from if frame doesn t include whole pop undercoverage will occur to some exten1 0 Ex survey about US adults many no reach deployed soldiers overseas o Bias due to taking voluntary response or convenience sample 0 Random sampling error variability NONSAMPLING ERROR errors not blocks caused by act of taking sample each block Can t be xed by changing sampling MalesK drug grougt Co re method Placebo ngup Can be present even in a census compare PROCESSING ERROR 7 someone makes a manual Femagdrug 9mm ompare error ex misrecording Placebo QFOU o POORLY WORDED QUESTIONS 0 RESPONSE ERROR occurs when subject gives incorrect response by lying remembering incorrectly not understanding the question etc MATCHED PAIRS DESIGN design to compare two different treatments 1 Pair up subjects ideally the two subjects are very similar to each other each subject in a pairing randomly receives one of the treatments blocks pairs 2 Each subject gets both treatments but in different orders ideally blocks each subject individually VALIDITY measure of a property that is relevant or appropriate as a representation of that property is the instrument the right choice for measuring the variable of interest RELIABILITY consistenczz measurement where the random error is small how consistent the instrument measures improve with average DOUBLEDOUBLE EXPERIMENTS neither the subjects nor the people who work with them know which treatment each subject receives eliminatesreduces bias MEAN the average of all data values MODE biggest peak most common number MEDIAN middle number in the data set STANDARD DEVIATION average distance of data values from the mean RANGE overall distance between min and max values m range of the middle 50 of the data the distance between Q1 25th percentile and Q3 75th percentile VARIANCE 1 11 1 2x1 mecm2 HISTOGRAMS quantitative variable distribution bars have equal width but height of each bar displays how many individuals fall within a speci c range of values UNIMODAL 1 peak BIMODAL 2 peaks MULTIMODAL more than 2 peaks BOXPLOTS GOOD skewedness vs symmetry shape center median and spread IQR comparisons between groups BOXPLOTS BAD number of peaks in a distribution size of data set frequency of values within different intervals OUTLIERS data points that deviate unusually far from the overall pattern circles separately from the lined part of box plot HOW TO FIND 1 Find Q1 and Q3 2 Calculate IQR Q3 Q1 middle 50 of data points 3 High outliers are gt Q3 15 x IQR 4 Low outliers are lt Q1 15 x IQR If distribution is skewed andor has outliers use measure of center measure of center median spread IQR If distribution is symmetric with no outliers use measure of center mean spread standard deviation RESISTANCE how strong of an impact outlier have on a particular measure of center or spread Mean and standard deviation are NOT resistant to outliers they re strongly impacted by outlier because use mean and standard deviation to calculated values for data Median and IQR are resistant to outliers no impacted by outliers because median is the middle value and same reason for the IQR
Are you sure you want to buy this material for
You're already Subscribed!
Looks like you've already subscribed to StudySoup, you won't need to purchase another subscription to get this material. To access this material simply click 'View Full Document'