Chapter 4.1 Vocab Flashcards
Convenience Sample
A convenience sample is choosing individuals from the population who are easy to reach. This causes data to be unrepresentative of the population because the sample is not carefully chosen with regulation that remove factors that may cause bias.
Bias
When factors in a statistical study cause constant over or under estimation of the value that on wants to know
Voluntary Response Sample
Individuals choose themselves to be apart of sample by responding to a general invitation. This can cause bias because those who have strong opinions, or share the same opinions are mostly likely to respond causing an overestimation of the sample statistic. Additionally, those who respond do not represent a specific population
Random Sampling
Leaving the Selection of the members of a population to be apart of a sample to chance. This can be done by picking from a hat, or using electronic methods of randomizing. Random Sampling ad SRS can be trusted because by using impersonal chance the methods remove bias.
SRS (Simple Random Sampling)
Methods of random sampling in which every individual of the population has the same chance of being selected to be apart of the sample. Some methods also voice every possible sample of the desired size an equal chance of being chosen.
Can be chosen in many ways one of which is randomizing manually, randomizing with technology, or randomizing with a table of random digits.
Table of Random Digits
Label your entire population with numbers and make sure they all have the same number of digits and use as less digits as possible. Then using a table of random digits to compare those numbers to those you labeled. On the list of random digits if the number exceeds your highest amount then skip the number, and if the number has already been recorded skip that as well. Do this until the desired sample amount is created.
Stratified random samples
When a population is classified into smaller groups of similar individuals called strata. Each strata must be different from each other. The strata should be chosen biased on given or previously known information. Used when the collecting a sample is too large/too much time/money, we collect data from smaller groups and combine the data to develop stratified random samples.
Cluster Sampling
In order to develop a cluster sample one must first place the population into groups of individuals who are located “near each other”. Each group is called a cluster and must be diverse in relation to the information because the cluster will accurately represent the entre populations variability. Then randomly choose a certain number of clusters and everyone in those clusters is apart of the sample. Cluster samples are used most often for practical reasons like saving money or time.
Multistage samples
When large scale samples are taken by combining two or more sampling methods.
Undercoverage
Occurs when not everyone in the population can be chosen into the sample
Nonresponse
When people who have already chosen in the sample can not be reached, or refuse o participate
What else can case bias?
Wording of the question and the order in which the questions are asked.
