Lecture One

Question 1

What is Simpson’s Paradox?

Answer 1

For quantitative data: a positive trend appears for two separate groups, but when the groups are combined there is a negative trend.

Question 2

1. What phenomena was behind the UC Berkely gender bias?
2. Describe what was happening analysis.
3. Why was this occurring?

Answer 2

1. Simpsons Paradox
2. Men who were applying for positions were significantly more likely to be admitted than female applicants overall at the university. However, when looking at individual faculties it seems as though there was a bias in favour of women and not men.
3. Rather than because of discrimination the statistics show that it was because women were applying for more competitive faculties with low rates of admission whereas men were applying to faculties that had a high rate of acceptance.

Question 3

What is the difference between disaggregated data and aggregated data?

Answer 3

Disaggregated data is numerical or non-numerical information that has been collected from multiple sources and/or multiple measures, variables or individuals.

Aggregate data is a combination of the disaggregated data into a summary of the data so it can be reported or used for statistical analysis and then broken down into component parts or smaller units of data.

Question 4

What is belief bias?

Answer 4

The tendency to judge the strength of arguments based on the plausibility of their conclusion rather than how strongly they support that conclusion.

Example:

No addictive things are inexpensive

Some cigarettes are inexpensive

Therefore, some addictive things are not cigarettes

The argument is invalid and implausible as cigarettes can encapsulate all addictive things, yet 71% endorsed it.

Question 5

What does IV and DV mean?

What is the difference between the two?

Answer 5

Independent Variable (IV) and Dependent Variable (DV)

In an experiment, the IV is changed to see how it affects something else, whereas the DV is a variable that is being measured/observed.

The IV is something I change and the DV depends on the IV

Question 6

What is the difference between Qualitative and Quantitative data?

Answer 6

Qualitative data is information about qualities; information that can’t actually be measured.

I.e Softness of your skin, the grace of which you run, colour of your eyes.

Quantitative data is information about quantities; that is, information that can be measured and written down with numbers.

I.e. Your height, your shoe size, length of fingernails.

Question 7

What is psychological measurement

What are the steps you need to take when measuring?

What are the two processes involved?

Answer 7

Scientific Investigation

Step 1: Theoretical Constructs

Step 2: Measures

Step 3: Data (observations)

The two processes involved are Operationalisation and Measurement

Question 8

What is a theoretical construct?

Answer 8

They are unobservable psychological entities.

E.g. Attitudes, beliefs, hopes

They are tools to help us make sense of ourselves and others

Question 9

What is a measure?

Answer 9

A measure is a tool for getting people to produce data.

E.g. Survey items, reaction times, blood oxygenation level

Ideally, it elicits data that are informative about the (theoretical) construct

Question 10

What is data and observations?

Answer 10

They are what you get when you use a measure.

E.g. Answers to questions, response speed, etc

These are things you can actually observe.

Question 11

What is Operationalisation? What does it relate?

Answer 11

Operationalisation relates ‘constructs’ to ‘measures’

E.g. ‘Beck Depression inventory’ tries to measure ‘depression’

‘Inspection time’ tries to measure ‘processing speed’

Operationalisation is the process of finding a good measure

Question 12

What does measurement relate to?

Answer 12

Measurement relates ‘measures’ to ‘observations’

E.g. Running an experiment, sending out the survey

It is the process of applying a measure to get the data (or observation)

Question 13

How many scales of measurement are there?

What are their names?

What order do they go in?

Answer 13

There are four scales of measurement.

Their names are NOIR (or IRON for easy out-of-order remembering):

Nominal​ (1st)

Ordinal​​ (2nd)

Interval​ (3rd)

Ratio​ (4th)

Question 14

What is a ‘​Nominal Scale’?

Answer 14

Nominal means ‘name’

This means the possible values have no particular relationship with each other.

There is no meaningful numbering scheme we can use.

E.g. 1. Blue eyes: 2. Brown eyes: 3. Green eyes;

VS 1. Green eyes: 2. Blue eyes: 3. Brown eyes

Neither scheme is any more meaningful than the other

1st order of measurement - named variables

Question 15

What is an Ordinal scale?

Answer 15

Ordinal scales are scales that have order, but there is no standard of measurement of differences.

I.e. 1. A tennis ladder is an ordinal scale since one can say that one person is better than another, but not by how much (for example you can’t say that Djokovic is 3.2% better than Federer).

2. Being Punched < Doing Statistics < Eating Pizza

The numbers are informative and ordered sequentially, but the magnitude of difference is unknown.

Ordinal means relating to the order of something in a series

2nd level of measurement - named + ordered variable

Question 16

What is an Interval scale?

Answer 16

Interval scale is a quantitative measurement scale where the difference between 2 variables is meaningful.

The interval equates to the interval (or distance) between 2 variables.

The numbers have a natural ordering, differences between numbers are meaningful, ratios between them are not.

I.E. 1902-1932 is a 30-year difference. 1902-1992 is a 90-year difference. Meaning the 1992 gap is 3x the size of the 30-year gap.

3rd order of measurement - named + ordered + proportionate interval between variables

Question 17

What is a Ratio Scale?

Answer 17

A ratio scale is a type of variable measurement scale which is quantitative in nature. Ratio scale allows any researcher to compare the intervals or differences.

The ratio scale possesses a zero point or character of origin. This is a unique feature of the ratio scale.

Numbers have a natural ordering and ‘zero means zero’ (0 is the character of origin). The differences between the numbers are meaningful. Ratios between them are also meaningful.

I.e. 20-40 vs 20-80. The graph starts at 0 years. 20-year vs 60-year difference. 3x the size of the small one. 40 is 50% of 80.

4th level or measurement - named + ordered + proportionate interval between variables + can accommodate absolute 0

Question 18

1. What are the two types of non-categorical variables?
2. What is the difference between them?
3. What scales of measurement can each use?

Answer 18

1. Discrete Variable and Continuous Variable
2. A continuous variable can take on any value between two specified values.

I.e. Weight measured between 70-100 kilos. Continuous because there are values in between (78.151, 91.7 etc.)

A discrete variable cannot take on any value between two specified values.

I.e. Counting number of heads on coin flips. It can be a number between 0 and infinite but it cannot be 4.5 flips.

3. Discrete can use all 4 scales of measurement, continuous can only use interval and ratio.

Question 19

What is the difference between Categorical and Non-Categorical variables?

Answer 19

Categorical variables are qualitative (names of labels, colours etc)

Non-Categorical variables are quantitative (numerical)

Question 20

In regards to scales of measurements; which ones are categorical, which ones are non-categorical and which ones are continuous variables and which ones are discrete variables?

Answer 20

Nominal = Categorical and discrete

Ordinal = Non-categorical and discrete

Interval = Non-categorical, continuous and discrete

Ratio = Non-categorical, continuous and discrete

Question 21

1. What is a ‘variable used to explain other variables’?
2. What is ‘variable to be explained in terms of other variables’?

Answer 21

1. It is a predictor, independent variable, treatment etc.
2. It is the outcome, dependent variable, response etc.

I.e. Smoking (predictor) causes Cancer (outcome)

Question 22

Do explanatory predictors have to be causal?

Can variables play multiple roles?

Answer 22

1. No. There can be factors in between.

E.g. Religious people live longer. Is it because of their religion? Not directly, but their religion makes them happier and then happiness gives them a longer life.

2. Yes.

In the example above, religiosity is associated with happiness, and happiness makes you live longer.

Question 23

In wide format; when analysing data what does the row and the column represent?

Answer 23

Each row represents a case (person)

Each column is a distinct variable

Question 24

What does cleaning data mean?

Answer 24

It means when your data isn’t input correctly and you need to make a decision whether to enter in their data correctly or decide if you have to get rid of the data/individual/variable altogether.