Please enable JavaScript.
Coggle requires JavaScript to display documents.
UNIT 4 project (Vocabulary in Experiments (Factors - influences on the…
UNIT 4 project
-
sampling error
convenience sampling
definition: It is when someone samples for something at their own ease. Sampling people or things that are easy to get to.
-
-
undercoverage
Definition: Often a problem with convenience samples, It is a type of selection bias and occurs when a part of a sample is not covered well enough. It is when some of the individuals in the population are not covered well enough in a sample.
example: surveying only people found at Trader Joe's would be undercoverage because the people found at Trader Joe's would be more likely upper-middle class, suburban, and white. It would leave out lower-income people.
non-sampling error
-
response bias
definition:Is also called survey bias, it is when people untruthfully or misleadingly answer questions on a survey.
example:If your coach asked you how many miles you ran and you say 10 miles, just to please them.
non-response
definition: non-response bias is when the poll is not compulsory, so a certain type of person is more likely to ignore the question
example: People that use illegal drugs would be less likely to answer questions on drugs in a survey.
Good Sampling Techniques
Stratified SRS
Advantages: better precision, smaller sample (saves money, quicker and easier to do)
Disadvantages-does not give the entire population a chance to get surveyed, cannot be applied in every situation
defintion: this kind of sample divides a population into different groups to ensure that responses are recorded from random selections within those groups
cluster sampling
Advantages-the entire population can possibly be surveyed. Only certain chosen groups of the population are surveyed. So this is cheaper and quicker than surveying the entire population.
-
definition: in this type of sample, the population is divided into different clusters– a simple random sample of clusters is selected and analysis is done on selected clusters
SRS
-
-
Definition- It is a group of individuals from a subset of a larger group. All of these individuals have an equal chance to be chosen from the whole set.
Experimental Design
-
Randomization-This is important to make sure the experiment or study gives a correct result. If there is minimal randomization then there may be bias in the survey.
Control: minimizes effects of other variables (besides independent variable) to make experiment better
-
Experiments
Observational Studies-These are conducted when experiments are non-ethical or not possible. Usually include surveys. Can help draw a conclusion to a certain topic for a sample population.
Example: an observation is done where they study the prevalence of marijuana usage and criminal activity
-
-
-
-
Example Experiment
Design Map
Experiment: 40 fish are randomly divided into two groups with 20 fish in each, both are put in separate tanks. Group 1 is the control group and is fed normal fish food, Group 2 is the experimental group and they are fed a new special type of food designed to make fish live longer. There will be no issues with placebo effect because the fish are completely unaware that an experiment is going on.
-