Please enable JavaScript.
Coggle requires JavaScript to display documents.
Producing Data (Sampling Error (Convenience Sample (Choosing the easiest…
Producing Data
Sampling Error
Convenience Sample
-
Example: We are surveying the most popular type of shoe brand, so we go to the mall and survey people in the shoe store.
-
Undercoverage
-
Example: A group of people that have different response for the survey might be accidentally left out of the sample causing a different result.
Non- Sampling
Response Bias
People that did not respond to a survey might be influence to say that they did respond to the survey.
Example: Someone that did not respond to a survey might be talking to the creator of the survey, and they might be feel as if they need to say that they answered the survey.
Non-Response
-
Example: I sent out an email to 1,000 people asking about their satisfaction about their cable provider. The non-response people are the ones that do not respond to the survey.
Wording Bias
-
Example: Due to the recent bear attacks, do you think it would be right to put the new bear laws in place.
Bias
-
Direction
-
Example: A coverage bias may only survey a group of the population and an under coverage bias may only use people who are easy to reach.
Definition
-
Example: A survey is only sent to a certain group of people so that the surveyor gets their desired result.
-
-
Experimental vocabulary
Response Variable: The response variable is the result of what happens from the treatments. The variable can change from treatment to treatment.
Treatments: Treatments are conditions that are applied to the individuals. A treatment is the combination of more than one explanatory variable.
Factors: Factors are the explanatory variables in an experiment. Experiments often use more than one factors.
Experiments
An experiment tests a hypothesis. The independent variables in a experiment subjected to some sort of treatment to get the desired outcome.
-
Observational Studies
An observational study takes the stats from a population. The study is not in the control of whoever is doing it because of ethical concerns.
-
Placebo
Blinding
The subject of the experiment is blinded on what type treatment they are given. They will not know if they received the placebo or the real treatment.
Placebo
Placebo is when a subject is given something that does nothing but is told that it does something. The subject will think that the placebo is effecting them but it does nothing.
Lurking Variable
Lurking variables lead to confounding when the lurking variable is associated with the explanatory variable.
Example: Does ice cream consumed increase the amount of drownings? The lurking variable is the temperature.
Survey
We are going to survey 10 different public schools on if they use a 4.0 GPA scale or any other GPA scale. To choose the schools we will randomly select 25 different schools then order them alphabetically and assign them a number. We will then use a random number generator to decide what 10 schools we will survey.We will not allow repeats of numbers, and if the number is over 25 then we will discard it and move on to the next number. After we have have the 10 schools we will carry out the survey.
This experiment used a 1.20 acre plot of land that was only dirt in order to test two types of grass seed side by side. The land was broken up into 4 .30 acre sections in order to make sure that both of the seeds had the same amount of space.
This created 3 different treatments for the grass seeds. The control group of the experiment was a section of the
land that did not receive any of the grass seed.
In the experiment the response variable was the height of the grass seed after growing for a month. Replication is used in this experiment by showing that a certain type of grass seed can grow in different sections of a field. Randomization is shown by having the two types of grass seed in every one of the four sections.
-
