Experimental Design
(After you observed, asked questions, and made a hypothesis.)
DEFINE A POPULATION
POPULATION: What is the group you’re trying to talk about? (recommended pop. is 200) The bigger the population, the far more difficult and costly, but more meaningful the statement becomes. The smaller the population, the less meaningful. It’s quick, dirty, and less expensive.
QUESTION: What exactly are you trying to say? You can’t say it’s true of the world. (EX: If you only sample Sanchez, you can’t assume the world is the same way.)
RANDOM SAMPLE: Draw random sample 1 & 2 (Experiment group, Control Group)
Samples must be diverse. There are individual differences.
200 or more people is generally a good sample size (valid)
20 people is the minimum number of people
INDEPENDENT VARIABLE: Give treatment A (Exp) Withhold Treatment A (Con)
Six weeks is necessary to show training effect due to Physiological adaptation. (min 6 wks.)
psychosomatic/placebo effect.
DEPENDENT VARIABLE Measure subject Measure subject
behavior on TASK Q behavior on TASK Q
DESCRIPTIVE STATISTICS: Calculate average of each group.
INFERENTIAL STATISTICS Compare XE with XC
Determine if the difference is significant.
Judge whether there truly is a difference. It’s entirely mathematical. Is it a significant difference or not? They take descriptive stats and compile them together. Psychosomatic effect (mind) – "Psycho" means mind. "Somatic" means body. "The mind can heal."
Remember
Placebo effect – Give false drug to make people think that they have the good stuff. They work just as hard as the experimental group. (Double Blind – neither knows that they have it.)
In science, always do the best you possibly can, then argue.
What is the effect of Y on X?
Choose something easily measurable: EX: Heart rate, strength, vision, body weight, muscle mass, fine motor skills (reflex), oxygen consumption in volume, etc.
Experimental Design Sample 1
Give the treatment or the whole treatment. Test each of the groups. EX: Take weight. Then take average. (Calculate mean).
TEST: Measure effect of Creatine Phosphate – increase strength.
25 lbs. vs. 10 lbs.??????
25 lbs. vs. 15 lbs.??????
(See Numeracy)
Experimental Design Sample 2
Trying to test if Chorpyrifos is present in the grass of a golf course in which a trustworthy source says it’s true. You’re trying to prove it.CHORPYRIFOS – Kills insects. ORGANOPHOSPHATE
PESTICIDE – Golf Course. It kills a lot of bugs to create a lot of grass.
CHOLINESTERASE – Chemical which causes your neurons to fire.
Cholinesterase inhibitor – nerve gas stops nerves from operating.
The muscles, especially the diaphragm, are the first to be affected.
Suppose you measured the soil, 10 to 15 feet in the morning and night. You did this for every few months. The water DID not wash of the chorpyrifos because the test proved negative when you tested the rivers, the aquifers, the estuary, etc. The air (evaporation) of water could not have been a factor. The question of biodegration is good, but there is STILL no chorpyrifos.
Hence, you use a control in which you KNOW that another golf course (golf course B) does indeed have chorpyrifos. However, when you tested it, it was still negative. WHAT’S WRONG?
It must be a problem with the materials or method.
PROBLEM: Something happens with the beaker. The meniscus effect takes place in which the polar covalent molecules stick to the side of the beaker. Thus, the chorpyrifos molecules were simply at the sides of the beaker, an area that was not tested.