Psychology 253
(Data Analysis in Psychology)
Chapter 1
Statistical Methods
i. Descriptive Statistics (p.4)
ii. Inferential Statistics (p.4)
Measurement Scales
a. Nominal Scale (p.6)
b. Ordinal Scale (p.7)
c. Interval Scale (p.7)
d. Ratio Scale (p.7)
Continuous Scale vs Discrete Scale (p.8)
- For all above: be able to differentiate and to match scale/method with a scenario – see
tuts!
Chapter 2
- Distribution
o Positively skewed – skewed to left – more low scores
o Negatively skewed – skewed to right – more high scores
o Symmetric distribution – no skew
Possible Q&A:
A normal distribution: symmetrical, bell-shaped, can be used to calculate probabilities
The standard deviation of intelligence score of male students is 14,5 and that of
females is 18,5. This means male students are more homogeneous than female
students regarding intelligence.’
If Z = 1, then the population is smaller than the raw score.
A value like the standard deviation that is calculated for a population is the square of
the variance.
Advantage of mean over median: more stable than the median.
Positively skewed distribution scores that occur the least will be to the right side of
the distribution.
More likely to reject null hypothesis if the rest statistic we compute is bigger than .05.
In a study on the effectiveness of the different therapeutic techniques two judges
ranked twenty patients with regard to the regard of insight shown after the therapeutic
sessions. It is an ordinal scale.
P > 0,05 means the null hypothesis should be rejected.
If behaviour problems scores are normally distributed and we want to make inferences
regarding the probability that specific valves can occur we must calculate the Z
values.
The null hypothesis is rejected if p > .05.
(Data Analysis in Psychology)
Chapter 1
Statistical Methods
i. Descriptive Statistics (p.4)
ii. Inferential Statistics (p.4)
Measurement Scales
a. Nominal Scale (p.6)
b. Ordinal Scale (p.7)
c. Interval Scale (p.7)
d. Ratio Scale (p.7)
Continuous Scale vs Discrete Scale (p.8)
- For all above: be able to differentiate and to match scale/method with a scenario – see
tuts!
Chapter 2
- Distribution
o Positively skewed – skewed to left – more low scores
o Negatively skewed – skewed to right – more high scores
o Symmetric distribution – no skew
Possible Q&A:
A normal distribution: symmetrical, bell-shaped, can be used to calculate probabilities
The standard deviation of intelligence score of male students is 14,5 and that of
females is 18,5. This means male students are more homogeneous than female
students regarding intelligence.’
If Z = 1, then the population is smaller than the raw score.
A value like the standard deviation that is calculated for a population is the square of
the variance.
Advantage of mean over median: more stable than the median.
Positively skewed distribution scores that occur the least will be to the right side of
the distribution.
More likely to reject null hypothesis if the rest statistic we compute is bigger than .05.
In a study on the effectiveness of the different therapeutic techniques two judges
ranked twenty patients with regard to the regard of insight shown after the therapeutic
sessions. It is an ordinal scale.
P > 0,05 means the null hypothesis should be rejected.
If behaviour problems scores are normally distributed and we want to make inferences
regarding the probability that specific valves can occur we must calculate the Z
values.
The null hypothesis is rejected if p > .05.
