Carl Christian Caduhada

Psychological Statistics 1
University of St. La Salle

MEASURES OF CENTRAL TENDENCY AND VARIABILITY Representative of the data set, value which
most of the data points are found.
SUMMARIZING DATA
ARITHMETIC MEAN
Using statistical tools and procedures
appropriate for answering a research Sum of all data values and dividing the sum
problem or objective. by the number of observations. When the
total number of observations (n) is small,
MEASURE
the mean is very sensitive to extreme
Numerical representation of a particular values.
characteristic (variable of the study) of the
POPULATION AND SAMPLE MEAN (ARITHMETIC MEAN)
group.

PARAMETER
Measure calculated from the population
usually represented by letters of the Greek
alphabet.

STATISTICS
Measure calculated from the sample
usually represented by letters of the
English alphabet.

SUMMARIES OF QUALITATIVE DATA

For example, sex a qualitative data may be
coded with 0 (male) and 1 (female).
Population is symbolized by μ (mu).
Although there is a number used it does not
entail any numerical values and could not Sample is symbolized by 𝑥̅ (x-bar).
be applied with arithmetic operations.
Differentiation whether the mean should be
PROPORTIONS population or sample plays a key role in the
interpretation of the results in inferential
Ex. Two out of ten are males. research.
PERCENTAGES Pythagorean MEAN
Ex. Twenty percent of the students are GEOMETRIC MEAN
male.
Formula:
SUMMARIES OF QUANTITATIVE DATA
𝑛
√𝑥1 × 𝑥2 × 𝑥3 … 𝑥𝑛
Summarized in terms of the center and
spread of the distribution. The center of the Harmonic mean
distribution can be identified using an
appropriate measure of central tendency or Formula:
location. 𝒏
𝟏
MEASURES OF CENTRAL TENDENCY ∑
𝒙

1
Carl Christian Caduhada
Psychological Statistics 1
University of St. La Salle

WEIGHTED MEAN MODE

Value in the data set with the highest
frequency or occurs most often.

This may be non-existent if all the values

occur with the same frequency.

May not be unique if it exists where a data

set may have more than one value having
To get wX multiply X (%) to w.
the same frequency.
Each value is weighted according to its
Correlates to peak/ peaks in a frequency
importance in the overall group.
distribution.
MEDIAN Uniform (flat or no peak), Unimodal (one
Center of an array dividing it into two equal peak), Bimodal (two peaks), Trimodal
parts. It is the midpoint of the numbers (three peaks).
arranged in ascending order. Data must at
RELATIONSHIP BETWEEN THE MEAN AND MEDIAN
least be ordinal.
The shape of the distribution is important in
ODD TOTAL POPULATION/SAMPLE selecting the most appropriate measure of
The median is the middle number in the central tendency.
array. Symmetric Mean = Median Equal
EVEN TOTAL POPULATION/SAMPLE Positively Mean > Median Greater
Skewed than
The median are the two middle values in Negatively Media < Mean Less than
the array. The two numbers may be added Skewed
and divided by two.

OUTLIER MEASURES OF VARIABILITY

There is an extremely high value among Measures of dispersion or variability.
the given data set which lies very far from Numerically describes the degree of
the rest of the values. dispersion, scatter or spread of scores in a
Mean in consideration of an Outlier distribution.

The mean cannot be found at the center or Non-central positions

the same as the median. The median 𝑖
Quartiles: 𝑘 = ( ) 𝑛
remains the same. 4

𝑖
Skewed Distribution Representation. Deciles: 𝑘 = (10) 𝑛

The mean cannot be used to represent the 𝑖

Percentiles: 𝑘 = (100) 𝑛
distribution if the shape is skewed.

The median is a better representative of a INTER-QUARTILE RANGE (IQR)

skewed distribution since the extreme Formula: IQR= Q₃ - Q₁
value does not affect the median.

2
Carl Christian Caduhada
Psychological Statistics 1
University of St. La Salle

Quartile deviation Formula:

Formula: IQR⁄2

RANGE
Difference between the highest (maximum)
and the lowest (minimum) observation.
If SK > 0; positively skewed
Formula: Maximum Value- Minimum If SK < 0; negatively skewed
Value If SK = 0; symmetric
STANDARD DEVIATION  <-1 or >+1; highly skewed
 Between -1 and -½ or between +½
Difference between the value of each and +1; moderately skewed
observation (xi) and the mean. Most stable  Between -½ and +-½;
and most commonly used. approximately symmetric
Formula:
COEFFICIENT OF VARIATION
Relative measure of variation. Where the
SD is a percentage of the mean.

Greater the CV (coefficient of variation) the

greater the variability.

Allows the determination of volatility in

comparison to the amount of return in
business statistics.
𝒔𝒅
VARIANCE Formula: 𝑪𝑽 = x 𝟏𝟎𝟎%
𝒙̅

Square of the standard deviation. STANDARD SCORE

Formula: Z-score or transformed raw score
measuring the relative position of one
particular raw score compared to the whole
data distribution. (student: class, company:
businesses)

Mean is zero.
𝒙−x̅
Formula: 𝒛 = 𝒔

Positive standard score: z-score is above

PEARSONIAN SKEWNESS
the average.
Measure of relative asymmetry by
Negative standard score: given raw score
comparing shapes of two or more
is below or lower than the mean.
distributions.

3
Carl Christian Caduhada
Psychological Statistics 1
University of St. La Salle

PROBABILITY Impossible Event (∅)

Measures the likelihood that an event will An event which will never happen. An event
occur. It is the basis of inferential statistics with no possible outcome.
in making decisions under conditions of
uncertainty. SURE EVENT (S)

Provides a means for measuring the An event which will really happen. An
uncertainty of decisions and inferences event which contains all the elements of
based on data from samples and the sample space.
experiments. Expressing probability
DEFINITION OF TERMINOLOGIES Smallest value
EXPERIMENT (E) Smallest value that probability statement
Any activity that generate data with the can have is 0
following components: (indicating the event is impossible)

a. action Largest value

b. instrument Largest value that probability statement
c. number of trials or repetitions can have is 1
Outcome (Sample Point) (indicating the event is certain to occur)

Each possible distinct result of an Definition

experiment. The probability of an event (A) denoted as
Sample Space (S) P(A) is the sum of the weights of all
outcomes in A.
Set of all possible outcomes of an
experiment. Represented by an uppercase 0 < P(A) < 1
letter S. P (Ø) = 0
P (S) = 1
EVENT (E) Complementation
Subset of the sample space; a collection of In a given experiment, an event must
outcomes satisfying a condition under occur or not occur.
which it will happened. May be represented
by any uppercase letter expect the letter S. A’ (nonoccurrence of event A), i.e. event A
does not occur, then we have
TYPES OF EVENTS
P(A) + P(A’)=1, then
Simple Event
P(A’) = 1-P(A)
An event which has only one possible
outcome. Therefore, P(A’) (A complement) denotes
the complement of A
Compound Event
THREE APPROACHES TO PROBABILITY
An event which has more than one
 Classical or A Priori Approach
possible outcome.
 Relative Frequency Approach
 Subjective Approach

4
Carl Christian Caduhada
Psychological Statistics 1
University of St. La Salle

CLASSICAL APPROACH  The occurrence of one event (A)

automatically precludes the occurrence
Arose from games of chance and depends
of the other event (B).
on the assumption that all outcomes are
 In symbols, A ∩ B
equally likely.
𝐧(𝐄) A B
𝐏(𝐄) =
𝐧(𝐒)

n(E)= number of outcomes favorable to E Non-Mutually exclusive events

n(S)= total number of possible outcomes When it is possible for two or more events
to occur together.
RELATIVE FREQUENCY/ empirical frequency APPROACH
 Does not indicate that such events
Determined on the basis of the proportion
must necessarily always occur jointly.
of times that a favorable outcome occurs in
 In symbols, A U B
a number of observations or experiments.

No prior assumption of equal likelihood is A B

involved.
𝐟𝐀 SET OPERATIONS
𝐏(𝐀) =
𝒏
Union
n= number of times an experiment is
repeated Symbol: A U B

fₐ= number of times (frequency) an event A Probability that (either) A or B will happen
is observed in n repetitions of the (or both)
experiment
If A and B are mutually exclusive:
Subjective / personalistic approach P(AUB)= P(A)+P(B)
Probability of an event is the degree of If A and B are non-mutually exclusive:
belief by an individual that the event will
occur, based on all evidence available to P(AUB)= P(A)+P(B)- P(A∩B)
the individual.

Used in assessments of political stability

and economic conditions, perhaps A B A B
concerning investment prospects or
currency fluctuations.

Mutually exclusive and non-mutually exclusive

events
Mutually exclusive events
Two or more events which cannot occur
together.

5
Carl Christian Caduhada
Psychological Statistics 1
University of St. La Salle

Intersection NOR
Symbol: A ∩ B Symbol: (A U B)’

Probability that (both) A and B will happen Formula: 1- P(AUB)

Probability that For mutually

neither A nor B will exclusive events:
A B
happen. A ∩ B= Ø
P(A ∩ B)= 0

COMPLEMENTATION A B
A∩B
Symbol: A’= not A

Formula: P(A’)= 1- P(A)

CONDITIONAL PROBABILITY
Probability that A’ will not happen Finding the probability of an event given
that some other event has already
occurred. Conditional probability restricts
A
the sample space.

Probability of event A, given event B:

set difference P(A ∩ B)
P(A|B) =
Symbol: A-B P(B)

Formula: P(A-B)= P(A)- P(A∩B) Probability of event B, given event A:

P(A ∩ B)
Probability that A but not B will happen, P(B|A) =
ONLY A will happen. P(A)

MULTIPLICATIVE rules
A B Used to find the probability of two or more
events that occur in a sequence.

P(A∩B) “and”. Keywords are both and

Symmetric difference and.
Symbol: A ⊕ B MULTIPLICATIVE rules: Dependent events
Formula: P(AUB)- P(A∩B) Given two non-mutually exclusive events A
Or: P(A) + P(B)- 2P(A∩B) and B: P(A∩B)= P(A) P(B|A)
P(A∩B)= P(A) P(B|A)
Probability that A or B will happen but not A and B are dependent events as A
both. Only A or Only B will happen. influences the probability of B in the first
equation and B influences the probability of
A in the second equation.
A B

6
Carl Christian Caduhada
Psychological Statistics 1
University of St. La Salle

MULTIPLICATIVE rules: INDependent events Standard Deviation: 𝜎 = √𝜎 2

Given two non-mutually exclusive events A MEAN AND VARIANCE OF Binomial VARIABLES
and B: P(A|B)= P(A)
Formulas:
P(B|A)= P(B)
A and B are independent events as the Mean: 𝜇 = 𝑛𝑝
outcome of event A given event B is only
equal to A. Variance: 𝜎 2 = 𝑛𝑝𝑞

PROBABILITY Distribution POISSON DISTRIBUTION AND POISSON PROCESS

Random Variable- a variable whose value Number of outcomes occurring during a

is determined by the outcome of a random given time interval or in a specified region.
experiment. (Denoted by X or Y) λ= average number of occurrences in an
P(X=x) probability of every value interval

∑ P(X=x)= 1 X= actual number of occurrences in that

interval

e= Euler’s Number (2.71828…)

BINOMIAL Distribution
𝑒 −𝜆 (𝜆) 𝑥
There are only two possible outcomes: 𝑝(𝑥; 𝜆) =
𝑥!
success or failure.

Bernoulli Process is where an experiment

consists of repeated trials, each with two
possible outcomes, success or failure.
The probability of success remains
constant from trial to trial.

p= probability of success

q= probability of failure

n= number of trials performed during an

experiment of getting x successes

x= probability of getting x successes out of

n trials

𝑃(𝑋 = 𝑥) = ₙCₓ ×𝑝 𝑥 × 𝑞𝑛−𝑥

where: p+q=1

MEAN AND VARIANCE OF RANDOM VARIABLES

Formulas:

Mean: 𝜇 = ∑[𝑥 × 𝑃(𝑋 = 𝑥)]

Variance: 𝜎 2 = ∑[(𝑥 − 𝜇)2 × 𝑃(𝑋 = 𝑥)

Alt. Formula: 𝜎 2 = ∑[𝑥 2 × 𝑃(𝑋) − 𝜇2 ]

7
Carl Christian Caduhada
Psychological Statistics 1
University of St. La Salle

NORMAL DISTRIBUTION PROPERTIES

It is a statistical fact applicable in real life The curve is symmetric about the mean μ.
that everything gravitates towards the
normal. The curve is unimodal. It has a unique
mode where the center of the curve lies.
Normality connotes averageness.
The curve is asymptotic to the horizontal
Most people and things are mediocre and axis.
embraced in the mean class.
The total area under the curve is equal to
Only a few are superior and inferior. 1.

DEFINITION APPROXIMATION OF BINOMIAL DISTRIBUTION

A continuous random variable X has a As n increases in size, it will become a
normal distribution with a mean μ (location normal curve.
parameter) and a standard deviation σ
CENTRAL LIMIT THEOREM
(scale parameter).
1 States that the sampling distribution of the
1 − (𝑥−𝜇)2
𝑓(𝑥) = 𝑒 2𝜎2 sample mean x̅ is approximately
𝜎√2𝜋
normally distributed with population
GRAPH mean μ from which the samples are drawn.

The graph of a normal distribution is Normal approximation for x̅ will generally

symmetric. be good if n > 30, thus n = 30 is a guideline
to use the central limit theorem.
The mean, median, and mode are all found
in the middle or center. This is guaranteed in any type of
distribution. (Whether symmetric or
Location is the mean and scale
skewed.)
parameter is the standard deviation.
PARAMETRIC TESTS
Assumes a normal distribution of values, or
a “bell-shaped curve.”

Also referred to as a Gaussian distribution.

Generally more powerful (requires

smaller sample size) than nonparametric
tests (nominal and ordinal data).

Normality is an important assumption in

parametric tests.

NONPARAMETRIC TESTS
Used in cases where parametric tests are
not appropriate.

It is about 95% as powerful as parametric

tests.

8
Carl Christian Caduhada
Psychological Statistics 1
University of St. La Salle

Cases such as:

Distribution is not normal (skewed).

Distribution is not known.

Sample size is too small (less than 30)

to assume a normal distribution.