Please enable JavaScript.
Coggle requires JavaScript to display documents.
Concept Map (Vocabulary (convenience sample is one of the main types of…
Concept Map
Vocabulary
convenience sample is one of the main types of non-probability sampling methods. A convenience sample is made up of people who are easy to reach.
A convenience sample example would be if somebody wanted to know how many people are signed up for prom and goes down to ask the breadboard during lunch. This is a convenience sample because the person is only asking a small group of seniors when the whole upper class is allowed to attend.
A voluntary sample is made up of people who self-select into the survey. Often, these folks have a strong interest in the main topic of the survey.
If a radio host stopped a justin beiber song from playing on the radio and asked the people to vote if they like him or not, the diehards would call in and say that they like justin bieber but this doesn’t prove anything it just proves that the callers care enough to call and vote for him.
An undercoverage is when a survey occurs when some members of the population are inadequately represented in the sample.
-
If you asked a low income are who they were voting for you might get different results and predictions on who would win the presidential race then if you asked a wealthy neighborhood.
Wording Bias is when the research question is worded or proposed in a specific way to get a certain response.
Given that Michael Jordan has won six championships to lebron's three, It is plausible to assume that Michael Jordan is better than Lebron James?
Response Bias is a large general term for users surveyed to respond inaccurately or without giving their complete honesty for many different reasons.
Such as when a survey asks how often a person drinks alcohol and they under cut the amount they consume, to appear more healthy.
-
Being biased is kind of lopsided too: a biased person favors one side or issue over another. While biased can just mean having a preference for one thing over another, it also is synonymous with "prejudiced," and that prejudice can be taken to the extreme.
-
Good sampling techniques
SRS is a type of random sample that completely assigns an equal probability for the total population of a group of people. An example of this would be if a college sampled their entire student body on their satisfaction of the education at their school. The advantages are these tend to provide a look at the whole picture getting everyone involved and letting them all have a voice. The Disadvantages are that these surveys are incredibly time consuming to complete having to sample an entire group.
SRS stratified is a form of complete sampling that assigns people to certain groups before the survey based on specific information. An example of this would be if the city of Boston choose to survey people of voting day on whether or not they believed in climate change while grouping them by a political party beforehand. The advantages are that these samples can quickly show a correlation between two things as if the groups show a connection in the answer in the survey.
Cluster Sampling, Cluster sample is when you divide a population into even groups and survey one of the groups and multiply the responses. This might not work because one of the groups might not have as many of a different aspect like age. If one of the groups ends up having more 18 year olds than 50 years olds golf might not be as popular of a sport as the other groups. However if the groups are even cluster sampling is an efficient way to survey large amounts of people. An example of a bad cluster sample would be if we were to survey a malls customers in a food court and divides the population into 4 sides of the food court. If the customers then voted on what food they were eating you might get more of one answer based on location.
-
An SRS Survey that would be a perfect example would be if you wanted to survey a hotel’s population for satisfaction out of 1-5, but you wanted to do it randomly and there were 5 floors. Surveying all these customers would be nearly impossible, so the idea is to select the first 200, numbers that fall under 6,000 on a simple random number survey. Only taking each number once so that the surveys authenticity stays intact.
The response would be gather the 1-5 satisfaction rating of each customer, and calculate the mean rating. Another thing that could be done with this information is by finding the mean by floor so you could see which floor has the guests most satisfied.
-
Observational Studies
Observational studies are studies “where researchers observe the effect of a risk factor, diagnostic test, treatment or other intervention without trying to change who is or isn't exposed to it.” Observational studies can’t ask you do do anything they just study the actions and the outcomes of them.
Experiments entail giving a job or task to the subjects to get the data. Experiments might tell someone to take a drug and the other subject to not take the drug vs observational study is when you study the actions and don’t differ them.
The big difference between experiments and Observational Studies would be that in an experiment investigators apply treatments to experimental units and observe the effect of the treatment. In an Observational study can observe the subjects without applying any treatments to them.
Placebo/ Blinding
Placebo Effect: Placebo effect is when you blind test. The subjects would think they are getting a genuine factor but it wouldn’t be affecting them physically but mentally. An example of this could be giving two groups of people a drug at different dosage level and see who had the better results. Although both of them had a fake drug the group with the “higher dosage” would say they had the better result even though it was the same drug.
Blinding: Blinding experiments is when you do an experiment without informing the subjects on what you are actually testing. This is done to eliminate the bias and get true and honest results.
Experiment Vocabulary:
Experimental units: the unit for statistical analysis. Experimental units can include tomato plants in an experiment showing how the light affects the growth of the plants.
Factors: explanatory variables. A factor could include the light in the experiment explained above
Treatments: treatments might be what stage of light the plant would get in the experiment. One treatment would be full light the other treatment being half-light.
Response Variable: The end result of the experiment. Referring to the tomato plant experiment response variable would be the end of plant height.
Lurking variables are variables outside of the explanatory or response variables that tend to falsely identify relationships. They lead to confounding by creating a fake idea of connection, which isn’t the case as it does not use the correct variables.
One example of this is if you were to do a study on whether or not wearing sunscreen helps prevent cancer and the lurking variable would be the strength of the sun based on the area the subjects live in or the amount of time spent in the sun.