Lecture 4 - Probability, Sampling and Distributions Flashcards

1
Q

Define probability theory

A

The branch of mathematics concerned with the study of random phenomena, i.e. chance.

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

Using the Gaussian equation, we can predict the value of y for any value of x from just the…

A

Mean and standard deviation

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

A positive skew moves the data peak of a normal distribution to the…

A

Left, and vice versa for a negative skew

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

Pearson’s coefficient of skew uses… to…

A

The difference between the mean and median… Measure the skew in terms of both magnitude and direction (positive or negative)

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

When data is positively skewed, i.e. the tail is on the right of the mean, mean>…

A

Median>mode

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

When data is negatively skewed i.e. tail is on the left of the mean, mean<…

A

Median<mode

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

What percentage of the sample are within 1 s.d. of the mean?

A

68%

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

What percentage of the sample is within 2 s.d.s of the mean?

A

About 95%

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

What percentage of the sample is within 3 s.d.s of the mean?

A

About 99.7%

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

Parametric tests assume that the mean and standard deviation…

A

Accurately represent the population distribution

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

Data can be transformed by…

A

Performing a mathematical operation (s) on all the values recorded

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

Data transformation is useful for…

A

+ reducing the impact of outliers/skew
+ standardisation, e.g. z scores are a consistent, universal unit
+ to remove non-linear effects
+ theoretical - using different measures to better understand the data
+ making the data normally distributed so that parametric tests can be used.

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

The z score is calculated by…

A

Taking the mean from each score and dividing the result by the standard deviation

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

The z score tells us…

A

How many standard deviations we are above or below the mean (0)

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

Sampling error is the difference…

A

Between the mean of each sample and the true mean of the population (and other sample means)

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

The standard error of a statistic tells us…

A

How much how much that statistic is likely to vary from one sample to another

17
Q

The standard error of the mean is…

A

A measure of how confident we are that we know the true population mean, calculated by dividing the standard deviation by the square root of the sample size.

18
Q

S.e.m. is dependent on…

A
  • The standard deviation of the population (variability of the original data) - smaller = more representative, higher confidence level
  • The number of data used to create the sample mean - larger sample size = more representative, higher confidence level
19
Q

Confidence intervals are often used as an alternative to…

A

S.e.m.

20
Q

A confidence interval of a certain percentage indicates…

A

The likelihood of that range containing the true mean

21
Q

Confidence intervals are computed by…

A

Multiplying the s.e.m. by the s.d. values you want it to be between, e.g. for a 95% c.i. multiply by +-1.96

22
Q

Error bars can be…

A

Either 95% c.i.s or s.e.m.s.

23
Q

The normal distribution, using various methods, allows us to determine:

A
  • the probability of a certain score or range of scores occurring
  • the probability that the population mean falls within a certain range
  • the probability that populations are different, e.g. through error bar differences