Statistical reasoning Flashcards Preview

PHIL105 > Statistical reasoning > Flashcards

Flashcards in Statistical reasoning Deck (25)
Loading flashcards...
1

In regard to B, if A had not occurred, B would not have occurred
Have to be a brother to be an uncle. What kind of condition is this?

Necessary condition

2

In regard to B, A is enough to guarantee that B occurs. What type of condition is this?

Sufficient condition

3

Statistical propositions

Propositions that present quantitative evidence about a category of things

4

total

the entire number of items in a class S or the sum of their values on a variable P

5

Ratio

the total expressed as a unit of another class (a total expressed in relative terms)

6

Frequency

the number or proportion of Ss that have some particular value on a variable P

7

Distribution

the number or proportion of Ss that have each of the values (P, Q, R) on some variable

8

mean

the sum of the values that the Ss have on a quantitative variable P, divided by the number of Ss

9

Median

The middle value of the Ss on a quantitative variable P

10

Absolute frequency

Tell actual number of Ss that are Ps

11

Relative frequency

Tell us proportion of Ss that are Ps

12

Proxy variable

variable that affects result but is not measured/directly relevant in data

13

Internal validity

certainty around cause and effect of members in study

14

External validity

can the generalisation hold for a wider group

15

Statistical generalisation

aiming for a representation of a larger sample (want quantitative diversity and use random sampling)

16

Can be 95% sure that our claim i will fall within this probability (e.g. 5% of 46 = likely to be within 41 and 51)

Margin of error

17

How you choose the sample has effect on how random it is (e.g. testing test scores from first 10 to arrive at class)

Selection bias

18

How you word questions has effect on randomness of sample (e.g. do you support war or intervention)

Testing biases

19

When random sampling consider:

Margin of error
Testing biases
Selection biases

20

Priming

expose to a stimulus influences a response to a subsequent stimulus without conscious guidance/intention

21

confounding variable

undesirably affects outcome as third unknown variable

22

Experimental vs observational

Can't control all variables so important to know difference

23

bigger the effect

more like there is a causation, rather than a correlation

24

Statistical correlations

Specific type of relationship between two variables e.g. average income and education years

25

Correlation does not automatically imply...

causation