Lecture 1 Flashcards

(31 cards)

1
Q

Population

A

Set of units we are interested in studying

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

Variable

A

Characteristic or property of an individual population unit

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

Census

A

Act of measuring characteristics on ALL UNITS IN A POPULATION

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

Sample

A

Subset (Part) of the population on which characteristics are measured

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

Some good statistical practices (good thins to do)

A

REPRESENTATIVE SAMPLE, ask good questions, make statements we have the data to support

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

Descriptive statistics (Def)

A

Used to describe observed phenomenons using quantitative tools. Summarize and describe world around us (quantitatively)

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

Inference (def) - (Statistical inference)

A

Generalizing a conclusion from a sample to a larger population

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

Experimental studies (data) generality (2 concepts)

A

One attribute of interest (THE TREATMENT) that we have control over and one attribute that we’re interested in observing (THE RESPONSE)

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

Key part of experimental study

A

We control who gets the treatment (so there is a control group or placebo)

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

Why control who gets treatment in experimental study and why have a control group

A

To see if it is really the treatment that caused the response

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

Observational studies (difference w/ experimental ?)

A

We do not control who gets the treatment

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

Confounding variable (def)

A

Variable that could be causing both the treatment and response variables

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

Problem w/ observational studies

A

We CAN NOT discount (ignorer/considerer negligeable) confounding variables or factors

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

Verb to use for an observational study’s results

A

We conclude association (not causation)

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

How to eliminate counfounding. What is this called

A

Making each subset of the population representative (equal probability of being selected for the sample) -
Random samples

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

Random samples ‘‘synonym’’

A

Mini-populations

17
Q

Selection bias (def)

A

A subset of the experimental units in population (a subset of the population/ a subset of populaton that we could measure variable on) has no chance of being selected

18
Q

Non-response bias

A

Inability to obtain data on all experimental units selected for the sample (some choose ‘‘don’t know’’ or ‘‘no answer’’)

19
Q

Measurement error and causes

A

Inaccuracies in the values of data recorded (poorly calibrated lab equipement, vague questions, measuring variable that represents large concept)

20
Q

Two ways of summarizing data

A

Numerically and graphically

21
Q

Graphical displays are good for … (2)

A

Building and guiding intuition, provide picture of data

22
Q

Numerical summaries are good for … (2)

A

Confirming intuition and giving concise impressions

23
Q

Quantitative data are …

A

numerical in nature (blood pressure, temperature, …)

24
Q

Qualitative data are …

A

Categorical in nature (Hair color, uni major, tumour vs no tumour)

25
Derived categories (explanation)
Qualitative data created quantitative data (ex : pass of fail, high or not high blood pressure, letter grades)
26
2 types of charts for qualitative data
Pie charts, Bar charts
27
What could be used tu numerically summarize qualitative data
Frequencies (how many have this and how many have that)/counts or Percentages/Proportions
28
What is the Simpson's paradox
Situation when a third, counfounding factor, changes the interpretation of the relationship between 2 other QUALITATIVE VARIABLES
29
Cause of Simpson's paradox
Imbalance in the distribution of the categories of the third category w/ respect of the first two (example of the majors
30
Misleading graphs (a couple conclusions)
Use all graph to show variations clearly, Pie chart : no angle of view, always view from top.
31
Frequencies vs proportions
Frequencies can present info in a way more transparent way. Proportions can lead to bad conclusions (like ex. of M/F acceptance at Berkeley)