Name my sampling method!
You are interested in determining the average hourly wage of a clerk at a chain store, such as 711. You find ten 711 stores within a 5 mile radius, randomly select one of them and perform a census on the employees' hourly wages. What type of sampling method is this?
Cluster sampling
A newscaster decides to survery residents in the Tri-State region regarding their decision as to whether or not allow taxpayer funds to support a new NFL stadium being built. This will lead to which type of bias?
Voluntary reponse bias
Both observational studies and experiments require randomization in order to be valid. However, the latter differ in that _______________ to one of two or more groups is implemented.
Assignment
Confounding
True or false: A control group and placebo group, qualitatively, are the exact same thing in any experiment.
False. While a control group and placebo group serve the same function of establishing a baseline from which to compare against other treatment groups, a placebo group effectively receives no treatment, while a control group can receive a treatment.
You are interested in how money the average Goldberg's Bagel patron spends at the shop nearby. You toss a die (1-6), and it happens to land on 2. You toss the same die again and it happens to land on 4. You decide to begin sampling the second customer followed by every fourth customer after that. This type of sampling method is known as?
Systematic random sampling.
A school counselor would like to estimate the proportion of students at her high school who vape or use nicotine products. She randomly selects 50 students and asks them the following question: "Have you vaped within the past week?" This sampling method will lead to:
Nonresponse bias
This necessary principle of experimental design ensures that a sound baseline is established so that proper comparisons in response between two or more groups can be made.
Control
A researcher found a strong positive association between ice cream sales and shark attacks at the Jersey Shore. She concluded that ice cream sales cause more shark attacks. Her spurious conclusion was due to not taking into account a __________ variable that drives both ice cream sales and shark attacks.
Lurking
True or false: If the researchers of an experiment are aware of which treatment groups their subjects reside in, then the experiment is not double-blind.
False. The researchers/data collectors of an experiment absolutely need to know which treatment groups their participants are in. It is the participants and administers of the treatments who should not know what treatment group they are in in order for the study to be double blind.
You would like to know, on average, how many hours a typical senior at EHS works after school. You have a population of 77 students and assign them each a number 1 through 77. After shuffling their numbers in a hat, you draw 12 students and record their responses. This method of sampling is known as?
Simple random sampling.
A public servant would like to know whether or not his constituents are supportive of a new initiative to bolster the town's public school system. He randomly selects 100 households and phones them, asking the following: "Would you be willing to support the town's school improvement program even if it means your property taxes will increase by 2%?" This sampling method will lead to:
Response bias
Although this experimental design mechanism is effectively the same thing as stratifying your sample (i.e., it reduces variability within a very heterogeneous population), there is a different vocabulary term used for describing the [optional] principle.
Blocking
True or false: A confounding variable and the bias it introduces can be effectively removed by properly redesigning the experiment.
True!
Consider this experiment. You, the researcher, wish to know the difference that fertilizer with a biological additive vs. the same fertilizer without the additive will have on the growth of tomato plants. Your population consists of two different species of tomato plants, A and B. You block by species and randomly assign to each species one of two treatments: control (fertilizer only) or treatment (fertilizer with additive). How many treatment groups in total do you have?
Two (2).
You are interested in determining the amount of time students at EHS spend on homework. Your population is all 500 students attending EHS. You decide to split your groups into two: 7th-8th graders and 9th-12th graders. You then perform a simple random sample of each group. What method of sampling did you use?
Stratified random sampling
You are gathering data for your Stats project and are interested in determining the average height of students at EHS. You stand by the school's main entrance from 7:30 to 7:50 and record the heights of the first 40 students who arrive. This is of course convenience sampling and is biased because which particular group of students is almost certainly going to be excluded?
Upper-classmen (SENIORS!)
Me: "I conducted my experiment my study and found statistically significant results! That's fancy speak for saying that my results were extremely unlikely to have occurred by chance alone... YEAH SCIENCE!!"
Not me: "That's great! And how many times did you perform your experiment under identical settings?"
Me: "Just the once!"
Not me: "You fool. That's wrong. You failed to ____________ your study. Otherwise, how can we be confident that your results aren't actually a product of random chance."
Replicate
True or false: A lurking variable and the bias it introduces cannot be removed or separated out in an experiment nor can it be eliminated through proper randomization.
Also true! A lurking variable that has not been accounted for will always create flaws in any experimental design.
True or false: a factor in an experiment is the same thing as the treatment.
False. A factor is the explanatory variable that is being manipulated and receiving the treatment at different levels.
You would like to know the average commute time of EHS teachers, so you await the front entrance and sample the first 15 teachers that arrive between 7:15 and 7:30. What method of sampling did you choose?
Convenience sampling
A local veterinarian would like to estimate the average age of pet dogs owned by Emerson residents. She obtains permission to view the list of all pets registered with the Emerson Borough goverment. Her sampling method will invariably suffer from:
Undercoverage bias
Subjects
An experiment can have both a confounding variable and a lurking variable at the same time.
True again!
True or false: Every experiment can be double-blinded, and blinding is one of the required four principles of every experimental design.
False. Oftentimes, an experiment can only be single-blinded (subjects only) or sometimes not at all. That does not automatically make it invalid.
Blinding however should always be done if possible.