Unit 3 Concept map
Sampling Error- A piece of data that cannot be correct due to the sample being taken not being reflected of the whole population.
Undercoverage- The sample is not adequately represented in a survey
Voluntary Response Variable- A sample that consists entirely of volunteers. However, these samples are always biased.
Convenience Sample- A convenience Sample is made up of a population that is easy to access.
Non-Sampling Error- A term used to describe systematic errors not related to an error in sampling
Non Response- People who are selected for the sample are unwilling or unable to participate in the survey
Response Bias- The tendency for someone to answer questions on a survey in a biased way
Wording Bias- The focus of the data trying to be gathered is biased in the way that it is worded. It is trying to get a desired response due to the words that are chosen to sway the sample population
Bias - The tendency to over estimate or underestimate the importance of one sample or population
Sources- Often when there is already a goal to the study, bias can happen. This is dangerous in a survey because the results should be independent and the threat of bias can disrupt the validity of the entire survey.
Samples and Surveys
Good Sampling Tecniques
Stratified SRS: First you divide the population into "strata". There can be any number of these. Then choose a simple random sample from each stratum. Combine those into the overall sample.
Cluster Sampling: the researcher divides the population into separate groups, called clusters. Then, a simple random sample of clusters is selected from the population.
SRS: A simple random sample is a subset of a statistical population in which each member of the subset has an equal probability of being chosen.
Advantage: An advantage of a stratified SRS is similar to a cluster however it is a more randomized group, taken from each subgroup to get a very accurate data set. So and advantage is that your data is not biased or skewed to one group it is very even.
Advantage: One advantage of having your population sample to be completely random is that there is little to no risk of bias or skewed data.
Advantage: You get an even more accurate data set because you are randomly sampling 2 parts of the population
Disadvantage: The data can be skewed because you can be taking data from one group and not spreading it out. For example you can just be picking people from one section of a stadium, but due to their position in the stands they could share commonalities like salaries. It is not the most accurate of tecniques
Disadvantage: If you are trying to find patterns in a specific group this tecnique would not be very useful because of how random it is, and how it selects samples from each inner group of a population.
Disadvantages: If you want to favor one group of a population in your experiment it is impossible because of the randomness
Example: An assembly room isn a high school is set up completely randomly of students. If you would want to find out how many people ate breakfast out of the assembly you could take a cluster of people in the front row and gather the data. You could also take a sample of the people in the bottom rows middle rows and upper rows, to get a more diverse group of people.
Observational studies vs Experiments
Experiments: In an experiment investigators apply
treatments to experimental units (people,
animals, plots of land, etc.) and then
proceed to observe the effect of the
treatments on the experimental units
Observational Studies: In an observational study investigators observe subjects and measure variables
of interest without assigning treatments to
the subjects. The treatment that each subject receives is determined beyond the control of the investigator.
Differences: Generally an experiment is more hands on while an observational study is more hands off. An experiment seeks a treatment during the experiment or as a result, however and observational study there is not treatment and it is purely for observing something, not changing the outcome.
Lurking Variable: A lurking variable is a variable that is not included as an explanatory or response variable in the analysis but can possibly affect the way we see relationships between variables. A lurking variable can falsely identify a strong relationship between variables or it can show little relationship while there is actually a strong relationship.
Experimental unit: The thing that is can be assigned, at random, to a treatment. Ex. an individual animal. The experimental unit is also the unit of statistical analysis
An explanatory variable is a certain type of independent variable. The difference between an independent variable and an explanatory variable is that a variable is explanatory when it isn't independent for certain.
Response Variable: Variables of interest in an experiment that are measured or observed. AKA dependent variables.
Treatment: In an experiment when the explanatory variable manipulated by the experimenter.
3 Principles to a good Experiment Design
Randomization: to ensure that this estimate is statistically valid. Randomization prevents the introduction of bias into the experiment and provides the link between the actual experiment and the statistical model that is under the data analysis.
Local control: to reduce experimental error by making the experiment more efficient. This can be achieved by making sure each treatment is uniform under standard conditions.
Replication: to provide an estimate of experimental error. In an experiment this could account for an inaccuracy in the result that was found. So it is important to let the public know an estimate to these possible errors in order for them to judge how valid or invalid the experiment is
Blinding: Typically used to describe the awareness of patients or the doctors as to if they are receiving the placebo or the actual treatment.
Placebo Effect: As before mentioned, a placebo is a fake pill or type of treatment that does not have any affect on the patient physically. However, after being asked how the patient feels they will often say that their symptoms have improved, even if they were prescribed the placebo. This is called the placebo effect.
Experiment Design Link Title
Independent Variable: Piano Practice
Replication: The initial skill level of the musicians
Dependent Variable: Correct Keys
Treatment: Hours Practied
Randomization: Each student in the class was randomly assigned to either group A or Group B