stat unit 1 Flashcards

(47 cards)

1
Q

Individual

A

smallest thing that will provide us with data/which data can be collected

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

Variable

A

any characteristic of interest of an individual

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

Categorical (qualitative) variable

A

a set of groups or categories

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

nominal level

categorical data

A

assosciated with words or names

ex. state of residency

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

ordinal level

categorical data

A

can be classified in a non-numeric order

ex. small, medium, large

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

Numerical (quantitative)

A

variable defined by a set of numbers

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

interval level

numerical

A

associated with a range of numbers, but 0 is not clearly defined

ex. temperature 0 degrees C = 32 degrees F

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

ratio level

numerical

A

can be compared between individuals, zero is clearly defined

Ex. $0 = absence of money

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

population

A

entire group of individuals

typically very large

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

Sample

A

portion of population that provides us with data

sample size (N) less than or equal to popoulation

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

Parameter

A

fixed unknown number describing some characteristic of the popoulation (fixed - stays the same bc population doesnt change, unknown - population is too big to find exact value)

characteristic examples: average, median, std. dev.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

statistic

A

a varying and known number describing a characteristic of the sample
- varies with each sample taken from the same population
- used to estimate the perameter

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

observational study

A

only observe how individuals react within situations the individuals are already in
- researchers have no control over situations
- correlation

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

types of observational studies

A

surveys and ethical dilemmas
- describe a group or situation by passively collecting data

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

experiments

A

deliberately impose situations on individuals in order to see what happens
- describe a group or situation by actively collecting data
- cause and effect

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

Cause and Effect

A

Because of the situaton (imposed by researchers), something happened

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
17
Q

Correlation

A

there may be a realtionship between the two, but we cant be certain

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
18
Q

Bad Sampling Techniques (BST)

Voluntary Response Sample

A

Individuals participate only because they really care about the topic

Course evaluations, ppl who either love or hate the class participate

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
19
Q

Bad Sampling Techniques (BST)

Convenience Sample

A

Individuals participate bc they are easy to contact

20
Q

Simple Random Sample (SRS)

A

Eliminates sampling bias, like a lottery
- N individuals chosen from the popoulation such that each individual of the populaiton has the same chance of being chosen
- each possible sample size N has the same chance of being chosen

Use table A for problems, 1-9

21
Q

Stratified Sample

good

A

Divide population into groups (strata) such that all individuals in a group have something in common
- take SRS from within each group
- Ex. pop. = clemson students, strata = freshmen, sophomore, etc, sample = randomly select 500 students from each group
- “some” OR “from each”

22
Q

Cluster Sample

good

A

Divide population into groups (clusters) such that all individuals in a group have something in common
- take SRS of groups, sample is all individuals within chosen groups
- Ex. pop. = clemson students, clusters = residence halls (bc there are many), sample = randomly select 5 residence halls

23
Q

Systematic Sample

good

A
  • order population in some way
  • randomly select a starting individual and a k integer, such that k is greater than 1 and less than the population size
  • sample = starting infividual and every Kth individual after that until desired sample size it met
24
Q

Sampling Errors

A

Use of bad sampling techniques
- fix = change sampling to SRS (stratified, cluster or systematic)

25
Random Sampling Errors
Deviations between statistic and paramters - statistic is consistently smaller/larger than the parameter - fix = increase sample size
26
Non-Sampling Errors
Errors not caused by sampling technique - entire pop. isnt represented accurately = under coverage of a sub-population - erroneous or multiple inclusion = subpop. counted too many times - processing errors = mistake in mechanical tasks - response errors = individual gives incorrect respons - nonresponse = failure to obtain data
27
# observational studies Slanted question
questions can be worded in a way to force a particular response - biased
28
# observational studies confusing words question
words may not be confusing to you but could be to the reader - ex. football
29
2 questions in 1
two questions might not be obvious, which would lead to greater misinterpretation
30
31
double negatives
confusing to reader
32
properly ordered questions
if one question relies on or can be influenced by the answer from another, then they should be placed in approprate order
33
matching
specific observational study when ethics are involved -**reasearchers dont control anything**
34
# Experiments Response Variable (RV)
dependent variable - what we are hoping to see a change in - measures outcome or result
35
# experiments Explanatory Variavle (EV)
independent variable - explains or causes change in the response variable
36
# experiments Subject (S)
individuals studied in an experiment - subject = single thing studied in experiment - individual - single thing studied in an observational study
37
# experiments treatment (T)
any specidic experimental condition applied, by the researchers, to the subject - specific levels of the explanatory variable - EV are general, these are **specific**
38
# experiment Lurking Variable
Characteristic that affects the RV but is not included as an EV -** inevitable** -poorly designed experiment
39
Confounded
two variables when their effects on an RV cannot be distinguished from one another
40
Randomized Comparative Experiment (RCE)
designed in a way so that the treatments can be compared
41
Double-Blind Experiment (DBE)
Neither the subjects nor the experimenters know which treatment each subject is recieving
42
# principles of experimental design control
effects of lurking variables on the response variable by ensuring all subjects are affected by any lurking variables
43
# principles of experimental design Randomize
Assigment of subjects to the treatments to ensure the groups are as similar as possible
43
# principles of experimental design Use enough subjects
in each group to reduce chance variation in the results
44
# experimental designs Completely Randomized Design (CRD)
All subjects are randomly distributed to the treatments -hope to get the same number of subjects in each treatment group but not always possible
45
# experimental designs Matched Pairs Design (MPD)
put subjects into pairs so that they are as similar as possible - treatments are randomly assigned within each *pair * - only two treatments - sometimes each pair is a single subject who gets both treatments and the order of treatments is randomized
46
# experimental designs Block Design (BD)
groups of similar subjects are put into "blocks" - randomly assign treatments within each block - Conclusion made for each embedded CRD, not one conclusion combining the infor - based on most significant *unavoidable sources of variability* (gender, age, etc)