STAT 2004 Test 1 VT
Terms : Hide Images
| the art and science of learning from data | ||
| to be able to graphically report your data | ||
| the mathematics and the experimentation | ||
| using evidence that we collect to make conclusions about our question of interest | ||
| evidence collected | ||
| the group of items that the question is trying to inquire information from | ||
| a subset of the population | ||
| am example of a sample, but represents a population since the researcher is trying to collect as many subjects as one can from the population | ||
| involves computing statistical calculations and constructing graphical representations to summarize the sample data collected | ||
| involves using descriptive statistics from sample data to make formal conclusions to a research question involving its population with some level of certainty | ||
| a statistical value that describes the population | ||
| a statistical value that describes a sample taken from its population | ||
| published source well-designed experiment observational study survey computer simulation | ||
| - data already created - published in a periodical, electronically in a database, or available on the internet for public use - most popular and cheapest method - must acknowledge the source of data | ||
| - conducted by assigning subjects to experimental conditions (treatments) - process in assigning can be random or nonrandom - setting strict control over subjects and providing a treatment - minimizing the loopholes in your experiment | ||
| - only observing and recording data based on how subjects react in natural settings - there is no researcher- subject interaction - as a researcher it is important not to disclose to the subjects that they are being observed as it affects their behaivor | ||
| - using a questionnaire (either electronic or paper) where your subjects record responds | ||
| - on the spot with an interviewer via a computer - by an interview over the phone - on paper and mailed in - electronically by email or the internet | ||
| - models of a real-world situation - usually done prior to conducting an experiment - for private, government, or public funds, these are sometimes required to conduct real life experiment - used to save money, preserve the environment, or avoid dangerous risks | ||
| types- categorical and quantitative any characteristic recorded from the subjects of a study Ex. height, weight, gender, ethnicity, color, names, temperature, performance, levels of pain | ||
| characteristics that have responses that are classes, groups, or categories | ||
| characteristics that take on numeric values that convey relative magnitude of the variable | ||
| Nominal- Categorical Ordinal- Categorical Interval- Quantitative Ratio- Quantitative | ||
| names or numbers only to represent separate categories. No name or value is more significant than another name or value | ||
| names or numbers where only the comparison of "greater", "less", or "equal" are relevant. There is rank in names or values | ||
| considers relative order and the size of the difference between two measurements. Absolute zero does not exists | ||
| considers relative order, size of difference between two measurements, and the ratio of two values. Absolute zero does exists | ||
| Examples- collecting names of people; collecting majors | ||
| Examples- ranking a professor (1=poor, 4=excellent) | ||
| Examples- temperature | ||
| Examples- annual salary | ||
| different categories under each factor | ||
| particular combination of factors | ||
| baseline level where the other levels in that factor are compared to | ||
| the treatment with the combination of all the factor control groups | ||
| a possible factor that can be used to debate the validity of the cause effect relationship between the explanatory and response variables | ||
| when subjects don't know which group they have been assigned to | ||
| an experiment when data collectors and experimenters are also not aware of the subjects' assignments as they collect data | ||
| two treatments in which each subject is given both treatments (one at a time in no particular order). Object is to determine if primary treatment significantly had an affect on the response variable | ||
| time constraints ethical issues subjects costs | ||
| a percentage that tells how far off the estimated is from the true population percentage | ||
| bias that can occur from such things like nonrandom samples or under coverage | ||
| bias that can occur when your selected subjects cannot be reached, refuse to participate, or fail to answer some questions | ||
| bias that can occur when a subject gives an incorrect response (intentional) or the subject gives an incorrect response because the question is misleading (unintentional) | ||
| to select subjects from the population | ||
| list of possible subjects to sample from | ||
| the methodology used for selecting subjects from the sampling frame | ||
| to give subjects with certain characteristics a higher chance of being selected or participating in the study (Ex. ethnicity, income status, gender, age) | ||
| - simple random sample -R - cluster random sample -R - stratified random sample -R - systematic sample - convenience sample -NR - volunteer sample -NR - purposeful sample -NR | ||
| sample in which each possible sample has same chance of being selected | ||
| series of random numbers used to select individuals for a sample | ||
| sample in which the population is divided into a large dividing the population into a large number of clusters and selecting a random sample of clusters and using all subjects in the cluster for sample | ||
| sample in which population is divided into stratum (groups based on particular variable) and a simple random sample is collected from each stratum | ||
| sample in which researcher numbers each observation in the population 1-N; uses a random number table to select one of the observations; chooses a seed number (k) between 1- and N-1; chooses every Kth observation from the first observation sketched | ||
| sample that is easy to conduct. most commonly used for surveys than observational studies and well designed experiments. unlikely to represent population. may cause severe biasness due to time, location and participation | ||
| sample in which people are given complete freedom to participate. experimenter may not be guaranteed a large enough sample, sample may not be representative of population, high chance of bias | ||
| sample in which individuals are selected based on certain criteria; used most in qualitative research where population is small; must justify in the research the reason for the purposeful sample. |
