Stats 121 notes week 5
Stats 121 notes week 5 STAT 121
Popular in Principles of Statistics
verified elite notetaker
Popular in Department
This 5 page Class Notes was uploaded by Sydney Clark on Friday September 30, 2016. The Class Notes belongs to STAT 121 at Brigham Young University taught by Dr. Christopher Reese in Winter 2016. Since its upload, it has received 2 views.
Reviews for Stats 121 notes week 5
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: 09/30/16
Stats 121 notes week 5 *IMPORTANT* *EXAMPLE TEST QUESTIONS* Collection Data: Sampling We care about the population mean and standard deviation Characteristics about the population is called parameter o Parameters are to population as statistics are to sampling Sample survey o Definition: a type of observational study in which individuals report variables’ values themselves, frequently by giving their opinions o Purpose: to be able to use that sub set to make inference on a population Why sample? o Compared to census we use probability because: Practical Cheap Often more accurate Vocab o Population: entire group of individuals of interest o Sample: individuals that are selected from the population and measure o Parameter: numerical fact about the variable in the population o Statistics: corresponding numerical fact in the sample For sampling to work: o Explicitly describe population o Explicitly describe variable Bad ways to sample o 1.) convenience sampling Select individuals in easiest possible way o 2.) Volunteer response sampling Individuals select themselves o 3.) quota sampling Force the sample to meet specifies quotas Participants within a subgroup are not selected randomly, rather by convenience or some sort of judgment call o Why bad?? Bias Sample favors certain outcomes Impossible to assess uncertainty Probability sampling designs o Simple random sampling o Stratified sampling o Multistaged sampling Simple random sampling (SRS) o Take your population o Number population from 1N Population size=N o Use random digit and pull out how many you want to get o The sample size =n o Sample of specified size chosen such that every possible sample set of that size has an equal chance of being selected for the sample o Assign a number to each individual in the population o Use a random device to select desired number of individuals Stratified sample o Quota sampling done right! o 1.) divide people into groups o 2.) individuals within a group share a similar characteristic o 3.) select SRS from every group o 4.) Combine SRS’s Multistage sample o Most populations have hierarchical structures: o o Take sample at each level o 1.) 1. SRS of states o 2.) for selected states, SRS’s of counties o 3.) for selected counties, SRS’s of people o 4.) combine SRS’s of people Non sampling bias o Probability samples may still have bias due to: Undercoverage Nonresponse Misleading response Interview Question Order o Probability sample= Good samples Undercoverage o Some individuals have no possibility of being selected Ex: homeless, phoneless, missionaries Nonresponse o Selected individuals refuse to answer or can’t be contacted Examples: Hang ups, on vacation, refusal to mail census form Misleading response o Selected individuals lie or give inaccurate answers Examples: Have you cheated? Did you wash your hands? Are you a voter? o Sensitive questions are often subject to misleading response. Interview o Interviewer influences responses Examples: Rude, intimidating, subtle clues or gestures Question order o Order of questions promotes certain responses Examples: happiness question precedes debt question and vice versa o Open vs closed questions open questions are less restrictive, but responses are more difficult to summarize closed questions may be biased by the options provided closed questions should permit options such as “Other ” and/or “Not Sure” if those options apply o Question wording Can be misleading or confusing Example: loaded words, double negatives, wordy questions Observational studies o Subjects choose which treatment to receive or naturally belong to one of the treatment groups o Sample surveys are observational studies and with observational variables you can’t conclude cause and effect variables…. You can only say they are related to each other Experiment o Definition: Any study where you impose treatments on subjects o Purpose: determine whether or not something causes an effect o Goal: less confounded lurking variables and can validly draw causeeffect conclusion o Subject: individual to which a condition or treatment is applied o Factor: another name for explanatory variable Label over treatments o Treatment: one experimental condition o Response variable: characteristic measured on each subject Categorical response o Lurking variables: affect response variable but not planned factors o Control: an effort to reduce effects of lurking variables o Confounding: situation in which effects of lurking variables cannot be distinguished from effects of factors o Principles of Valid Experiments o 1.) Control/Comparison o 2.) Randomization o 3.) Replication o 4.) Double blinding Control/ Comparison o control lurking variables by including comparison treatments, using homogeneous subjects; used to measure placebo effect Randomization o neutralize effects of lurking variables by assigning subjects to treatments randomly Replication o assign enough subjects (> 1!) to treatments to be able to detect important effects Double Blinding o neither the subjects nor the people who evaluate them know which treatment each subject is receiving; used to prevent experimenter effect Valid experimental designs o Randomized Control experiment (RCE) Every subject has an equal chance of being assigned (names in a hat) o Randomized block design (RBD) Even if you do a good experiment you can still make mistakes o Randomized comparative experiments may still have problems placebo effect diagnostic bias lack of realism Hawthorne effect noncompliance Placebo effect o Response by human and nonhuman subjects due to the psychological effect of being treated o psychological effect is confounded lurking variable o Consequence: ineffective treatment appears effective relative to untreated subjects o Solution: 1. use dummy treatment (saline, sugar pill, etc.) rather than “no treatment” as comparison treatment 2. blind subjects as to which treatment they are receiving Diagnostic bias o Diagnosis of subjects biased by preconceived notions about effectiveness of treatment Preconception is confounded lurking variable o Solution: Randomization Lack of realism o Realism is often comprised by controlled study conditions, choice of subjects, application of treatments o Solution: Admit limitations of experiments Hawthorne effect o People behave different in an experiment than they behave outside an experiment Noncompliance o Failure to follow through with protocol of study Principles of data ethics o Safety of subjects must be protected o All individuals must give their informed consent before data are collected o Individual data must be kept confidential
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'