DECK 6 UNIT 3 PART B Flashcards
What is the sure way to assign treatments correctly?
throw names in hat and first half in group 1 and the rest group 2. . Or number subjects from 1-n and use randint until you get half for group 1.
How are we using random numbers in experiments vs studies?
In a study, we randomly choose subjects to survey from the population as a whole. In an expermint we randomly assign subjects to treatment groups
How are we making inferences differently in experiments vs studies?
In observaional studies, you make and inference about the population, in an experiment you make an inference about a treatment.
How are we manipulating the environment differently in experiments and studies?
No manipulation or treatments in an observational study. You only manipulate environment in an experminet.
How are we proving causation in experiments and obs studies?
No causation in a study, maybe association or correlation. ONLY EXPERIMENTS SHOW CAUSATION.
How do we use representative samples in experiments and studies?
You don’t need a representative sample in an experiment. You are not making inferences about a population, just about a treatment.
What is the difference between confounding and lurking?
Confounding is in experiments, like sunlight confounding a fertilizer experiment. Lurking is when you think hot chocolate causes ski accidents. “lurking”is actually a word not even used in AP STATS.
What is wrong with using voluteers in an experiment?
Not much. In an experiment, we are not looking for a sample that is like the population. We just want to see the effectiveness of a treatment. It is fine if the subjects are all similar. In fact it is best sometimes when they are!
What is a control group?
The group that doesn’t get a treatment (or gets the old treatment). It helps us see the impact of the environment. It gets the placebo or standard care but goes through all of the motions
Give example of confounding variable.
fertilizer A vs B. If you have two tables in a room with tomato plants and and one table gets A and the other gets B, but later you realize that the table with A was near the windows. You say that SUNLIGHT IS A CONFOUNDING VARIABLE in that experiment.
In the fertilizer experiment, how could you plan to eliminate the confounding variable?
USING RANDOMIZED BLOCK DESIGN. Make each table a block, and then randomly assign fertilzer A and B to the plants at each table. Compare the fertilizers for table 1, then compare the fertilizers on table 2.
Why does it make sense to double-blind an experiment?
It reduces bias in an experiment. If subjects don’t know what treatment they’re receiving, they won’t change their habits based on that knowledge. If evaluators don’t know which treatment each subject is receiving, they won’t bias the true results based on the results they expect to see
4 INGREDIENTS TO EXPERIMENTS
Compare, control , randomization, replication (and BLOCKING when you need to)
4 ingredients: What is “compare?”
Having something to compare your treatment with helps you see its effectiveness.
4 ingredients: What is “control?”
You want to control the environment as best as you can so that the only difference between groups is the treatment, and the treatment only. Everything else should be similar.
4 ingredients: What is “BLOCKING?”
If you think different groups of subjects may respond differently to treatments because of location, gender, age, then you make BLOCKS, and make sure to compare the treatments within each block.
4 ingredients: What is “replication?”
Having plenty of subjects in each treatment group. You don’t want to test fertilizer on just one plant.
4 ingredients: What is “randomization?”
You want to randomly assign subjects to treatment groups.
What are the two (three) types of experiments?
- Completely randomized 2. Randomized block (matched pairs)
What is the difference between a completely randomized and a randomized block?
In a completely randomized experiment, all of the subjects names go into ONE HAT and you pick for treatment groups. In a randomized block design you have more hats (a hat for males, a hat for females etc) and pick for treatments from each.
Give example of a matched pair design study for fuel efficiency
Testing fuel efficiency of different gasolines. Subjects use both fuel A for a month and fuel B for a month and compare, based on their driving habits and vehicle, which was more efficient. BE SURE TO RANDOMLY CHOOSE WHICH ONE GOES FIRST FOR EACH SUBJECT.
Give an example of matched pair design for comparing a new blood pressure medication to an older version?
Have subjects use the new medication for a month and the old one for a month and compare. Be sure to randomize which month and blind.
What is the placebo effect?
When those who get the placebo show improvements, or show the effects of the treatment. This often happens to up 20% of participants!
Who can be blinded? ( two groups)
- Subjects (and dog owners..). The poeple getting treatment and 2. administrators. Those delivering treatments and assessing effectiveness of treatments.
What is statistically significant?
When an observed difference is too odd for us to believe that it is likely to have occurred naturally (or just randomly). Basically it is Statistically Significant when we don’t think it happened randomly. when you think “something’s up” or “something’s fishy”
What is difference between subject and experimental unit?
Humans who are experimented on are commonly called subjects in an experiment. Subjects like dogs, days, plants and anything not human are called Experimental Units
What is the purpose of matching?
it isolates the differences between subjects so help see the impact of the treatment.
What “designs” are there for experiments?
completely randomized, random block, and matched pair design(type of block)
What is a factor? give example
DIET PLAN would be a factor and levels could be: low carb, low fat, and no diet
What do observational studies and experiments have in common?
In both, you are making OBSERVATIONS.. recording data… doing statistical analysis…
What’s a useful alternative when you can’t run an experiment? What are they useful forms of this, and how do you preform them respectively?
An alternative of an experiments could be an observational study. t.
Give example of randomized block design for a new anxiety med vs placebo with 100 volunteers (60m and 40f)
Block by gender. Randomly assign 30m to new and rest placebo. Randomly assign 20w new and rest placebo.
How is Blocking in an Experiment Similar to Stratefying in a Sample?
The two are similar because they divide the subjects into homogenous groups where the subjects are all similar (these traits were already present in the population)
What type of study would find relationship beween Verbal and Math SAT?
You could take all of the SAT Math and Verbal scores and run a regression and find the r-quared value and linear model. This would be a Retrospective Study.
What is the main purpose of a placebo ?
To blind the subject that is being experimented on to avoid influence to the given variable therefore altering the response variable . When people think they’re getting help, they often improve anyway..
What is matched pair design?
A type of blocking where you match subjects to other “like” subjects… MOST OFTEN SEEN WHEN YOU COMPARE A SUBJECT TO ITSELF!! (like pre-post tests)
Why randomize in an experiment?
To reduce confounding variables (and bias).
What is difference between completely randomized and random block design?
Completely randomized takes all units and puts them in a hat and randomly chooses treatments, blocked puts them all in different hats first (blocks) and then chooses
What is the difference between single-blind and double blind?
Single blinding is when all individuals in either one of the classes are blinded; double-blinded is when everyone in BOTH classes are blinded. Classes are: subjects and the other is treatment givers + evaluators…
Can you stratify in an experiment?
NO. stratification is a sampling method, blocking is method used in experiments. They are similar ideas.
name 2 differences between observational studies and experiments
- Experiments can prove causation (studies can’t) . 2. In experiments, you assign treatments (studies you just watch)
What is a level in an experiment? give example
Example. For the Factor “SLEEP” the, level(s) would be how many hours the subjects were alowed to sleep.. 4 hours, 6 hours, 8 hours.. 3 levels
Give example of factors and levels
Factor: medication. Levels: 50mg, 100mg and 200mg.
Suppose you are doing a weight loss experiment with 2 diets (A. low carb and B. low fat), and three medications (1. NUTRI LOSS, 2. POUND DROPPER and 3. SLIMMERLY). How many treatment groups would there be?
there would be 6: A1, A2, A3, B1, B2, B3
Suppose you are doing a weight loss experiment with 2 diets (A. low carb and B. low fat), and three medications (1. NUTRI LOSS, 2. POUND DROPPER and 3. SLIMMERLY). What is the response variable?
weight loss (pounds)
If you randomly choose subjects from the population what can you make an inference about?
You can make and inference about the population, you can talk about an association, but not about causation.
If you randomly assign subjects to treatment groups, what can you make an inference about?
You can make an inference about the treatment, about cause an effect.
