# ENVS 1030 Study Guide - Final Guide: Type I And Type Ii Errors, Null Hypothesis, Dependent And Independent Variables

by OC643495

School

University of GuelphDepartment

Environmental SciencesCourse Code

ENVS 1030Professor

Shelley HuntStudy Guide

FinalThis

**preview**shows pages 1-3. to view the full**15 pages of the document.**ENVS FINAL 12/17/2015

Scientic method

observation

hypothesis (general)

prediction (specic)

test (experimental observation

update probability

Observational vs manipulative studies

correlation vs causation

ocorrelation does not imply causation

relation to the problem does not imply it is the problem

directionality problem: don’t know which variable

is actually the cost

3rd variable problem: the problem might be caused

by a third factor/variable

Sample: is a subset of the entire group of interest

Random sample: is made by choosing in an unbiased way

by numbering the individuals and drawing numbers from a hat

Randomize

we want each unit to have an equal chance of being chosen for

oinclusion in the sample being measure (observational study)

oor, inclusion in a given treatment group (manipulative study)

oavoid BIAS

Only pages 1-3 are available for preview. Some parts have been intentionally blurred.

Type I error

the null hypothesis is true but you reject the null hypothesis (alpha)

Type II error

the null hypothesis is false but you fail to reject the null hypothesis

(beta)

Relationship between TYPE I AND II errors

decreasing the probability of type I error (alpha) increases the

probability of a type II error (beta)

for any xed alpha, an increase in the sample size n will cause a

decrease in beta

for any xed sample size n, a decrease in alpha will cause an

increase in beta

for any xed sample size n, an increase in alpha will cause a

decrease in beta

to decrease both alpha and beta, increase the sample size

Controls

negative: leave it exactly as it is (ex, give a group drugs while not to

others)

positive: placebo (ex, give some drugs others placebo)

historic: using old data, other control ideas

concurrent: make your own control

Blind procedures

a person measuring the dependent variable (s) is unaware of

treatment

addresses potential bias

Blocking

controlling for variables that might introduce more variation in your

response variable (ex, age of dog)

Split plot design

Only pages 1-3 are available for preview. Some parts have been intentionally blurred.

two factors

grouping by one factor

random allocation for second factor

convenience

Tradeo9 sampling

tradeo9 between sample size and information

Interactions (covariates)

the relationship between mating success and brightness of male

sh exterior

ointeractions or no interactions

Logical fallacy= ;awed reasoning

a claim or argument attempting to sway you without good evidence

oHasty generalization

a study shows that BPA is present in the urine of babies

who drink for plastic bottles “babies are being poisoned

by their bottles”

oRed Herring

“BPA may be harmful, but plastic companies employ a

lot of people”

oAd hominem attack

“BPA should be banned because the chemical industry is

untrustworthy and greedy)

oappeal to authority

“The national toxicology program is concerned about

the health e9ects of BPA, so it must be a health hazard”

oFalse dichotomy

“BPA must be completely avoided” “BPA is perfectly safe

for everyone” “Since BPA is harmful in some cases, it

must be avoided by everyone”

###### You're Reading a Preview

Unlock to view full version