PSYCHOLOGICAL STATISTICS

CHAPTER 1: INTRO TO PSYCH STATS

o E.g., t-test, ANOVA
Statistics
 Samples are imperfect representations of
 Statistics are a set of tools that help us to parameters

1. Organize data o Sampling error - the discrepancy

o Data - set of systematic between a sample statistic and a
measurements or observations population parameter
2. Summarize data
3. Interpret data Methods: Correlation
o Determine the relation between two
or more variables  The correlational method involves
measuring two (or more) variables to
 Variable – a characteristic determine whether there is a relationship
or condition of an object or between them
human that has different
values for different Methods: Experiments
individuals
 Experiments are a special type of research
o infer if differences exist between in which all the variables except the
different conditions in an independent and dependent variables are
experiment held constant

Population o Independent variable - the

variable that the researcher
 The population is the group of all people or systematically manipulates
objects that we are interested in o Dependent variable - the variable
that the researcher measures or
o Humans are often the population of records
interest in psychology studies
o Smaller populations are possible - all the Methods: Experiments
students in one particular class of
introductory statistics  Control condition the group that does not
receive the treatment
 Usually, the entire population cannot be  Treatment condition the group that
studied. Why? receives the treatment
 Population parameter – a value that  Performance in the treatment condition is
describes a population compared to performance in the control
condition to see if the treatment had an effect
Samples
Pick the Variable
 In most research, a subset of the population
called the sample is selected  A researcher randomly assigned participants
 If the sample is selected so that each to one of two groups - either the group that
member of the population has an equal consumes 3 beers or the group that
chance of being selected (called a random consumes no beer. Then both groups view a
sample), list of 20 words that they are to remember.
 and if the sample is sufficiently large, then Later they recall the words.
 whatever we learn about the sample will
probably also be true of the population o What is the independent variable?
 Sample statistic – a value that describes a o What is the dependent variable?
sample o What is the control condition?
o What is the treatment condition?
Descriptive Statistics
Methods: Experiments
 Descriptive statistics allow us to
summarize, organize and simplify data
 The main advantage of performing
 Measures of central tendency tell us about experiments is that they allow us, using
the average value statistics, to infer if the independent variables
cause the change in the dependent variable
o E.g., the mean, median, and mode

 Measures of dispersion tell us how similar Quasi-Experiment

the data are to the average value
 A quasi-experiment is similar to a real
o E.g., range, semi-interquartile experiment except that the participants have
range, and standard deviation been assigned to the various groups based
on some characteristics of the participant
Inferential Statistics
o E.g., participants are assigned to one of
 Inferential statistics allow us to study two groups based on whether they are
samples and then generalize about the male or female
population from which they were selected
PSYCH STATS CHAPTER 1: INTRO TO PSYCH STATS

 A researcher assigned students into one of larger than another added or subtracted
two groups - whether they over 8 years of multiplied or divided
age or 8 years or less. They were asked to  Cannot calculate the mean (what most
repeatedly recite the alphabet backwards people call the average)
until they did it perfectly. The number of
recitations of the alphabet was recorded. Ordinal Scale
 If there is a difference in the number of
recitations, can we say it is because of the  Ordinal scale consists of a set of categories
difference in age? organized in an ordered sequence. Spacing
between categories is not equal
Discrete and Continuous Variables
o E.g. Class rank, order of finishing a
 Variables can be either discrete (or horse race, how much you prefer
discontinuous) or continuous various vegetables
 A discrete variable consists of separate,
indivisible categories. No values can exist Ordinal Variables
between two neighboring categories
 The values of ordinal variables can be:
o E.g. number of children in a family compared to see if they are equal or not,
compared to see if one is larger or smaller
 A continuous variable can have any value than another
 The values of ordinal variables cannot be
o E.g. weight meaningfully: added or subtracted,
multiplied or divided
Continuous Variables  Cannot calculate the mean

 Measurements of continuous variables is Interval Scale

imperfect
 Real limits are the boundaries of intervals for  The interval scale consists of ordered
scores that are represented on a continuous categories that are all intervals of exactly the
number line same size. The 0 point is arbitrary.

o Lower real limit bottom of the interval o E.g. Shoe size, IQ scores, o
o Upper real limit top of the interval Farenheit

 Measure height to nearest cm: 182 cm Interval Variables

o Lower real limit is 181.5 cm
o Upper real limit is 182.5 cm  The values of interval variables can be:
compared to see if they are equal or not
Levels of Measurement compared to see if one is larger or smaller
than another added or subtracted
 When we observe and record a variable, it  The values of interval variables cannot be
has characteristics that influence the type of meaningfully: multiplied or divided
statistical analysis that we can perform on it o 70°F is not twice as hot as 35°F

o These characteristics are referred Ratio Scale

to as the level of measurement of
the variable  Ratio scale is an interval scale with an
absolute zero
 The first step in any statistical analysis is to  The absolute zero implies that a value of
determine the level of measurement; it tells zero represents the complete absence of the
us what statistical tests can and cannot be variable
performed
o E.g. Weight, reaction time, number
Nominal Scale correct

 The nominal scale consists of a set of Ratio Variables

categories that have different names in no
particular order  The values of ratio variables can be:
compared to see if they are equal or not,
o e.g. Hair color, whether a person compared to see if one is larger or smaller
has schizophrenia or not, five-digit than another, added or subtracted, multiplied
ZIP codes or divided
 Nominal variables deal with qualitative (and Representing Variables
not quantitative) differences
 By convention, in statistical formulae
Nominal Variables variables are represented by a capitalized
letter, usually X or Y
 The values of nominal variables can be
compared to see if they are equal or not o E.g., X might represent how
 The values of nominal variables cannot introverted the people in your
be meaningful: compared to see if one is sample are
PSYCH STATS CHAPTER 1: INTRO TO PSYCH STATS

Representing Individual Values

Ex.: Ex.:
 When a variable is subscripted (Xi), the
subscript implies that you should deal with a  N = number of  n = no. of
particular observation cases in a cases in a
population sample
o E.g. X3 might represent how  (mu) =  (x-bar) =
introverted the third person in your population symbol for
sample is mean sample mean
 σ (sigma) =  sd or s =
population std letter/s
The Summation of Operator (Σ)
deviation denoting
sample
standard
 Most statistical procedures involve the deviation
summation of the values of variables
 Rather than to write all the values out (X1 +
III. Sampling Errors and Sampling Variation
X2 + X3 + X4 + ...) a short hand notation is
used: ΣX
 Sampling error is the amount of estimation
 N represents the number of observations
error or discrepancy between a sample
statistic and the actual population value it is
i Xi supposed to estimate.
1 4  Sampling variation - variations in sample
2 2 statistics i.e., sample statistics (e.g., sample
3 -1 mean) obtained differ from one sample to the
4 7 next.
 Sampling error and sampling variation
ΣX = 12 occur even in representative samples.
N=4  Increasing the sample size reduces
sampling error and variation.
BASIC CONCEPTS IN STATISTICS

I. Preliminary Concepts IV. Statistical Model

a. Variables and Constant outcome = statistical estimate/s + error

b. Construct and Operational definition
c. Experiment and Correlation  An equation that reflects the relationship
d. Independent Variables / predictors and among variables.
dependent variables / outcomes  All models are false but some models are
e. Control group and experimental group more useful than others.
 Are usually variations of the equation above?
II. Samples and Populations
V. Types of Statistical Analyses
POPULATION SAMPLE
DESCRIPTIVE INFERENTIAL
A collection or set of A subset of a STATISTICS STATISTICS
individuals/units a population is selected Statistical procedures Statistical techniques
researcher is interested according to some used to summarize, that allow us to study
in. scheme, which is organize, and simplify and use the limited
intended to represent data information from
An arbitrarily defined the entire population. samples and make
group, the size and general probabilistic,
composition of which Findings from are conclusions regarding
rest on the researcher. presentative sample causes and/ or
may be generalized to relationships about the
A complete data set of the entire population. populations they
actual or potential represent.
observations. Mathematical Mathematical
procedures are used to procedures used to
PARAMETERS STATISTICS describe and conclude the population
summarize the from which the sample
Numerical values that Numerical values that characteristics of a was drawn.
refer to or describe a refer to or describe a group of scores or
population. sample. group of people in a "The purpose of
meaningful way. analyzing a sample
Derived from Derived from is to draw conclusions
measurements of measurements of "To make any about the population
individuals in the individuals in a sample. inferences about the from which the sample
population. population, we first was selected. This is
Usually used to need called INFERENCE and
May also be inferred estimate parameters. to describe the sample. is the primary goal of
from a sample. This is the primary goal inferential statistics."
of descriptive statistics."
Types of descriptive Types of inferential
statistics: statistics:
PSYCH STATS CHAPTER 1: INTRO TO PSYCH STATS

a. Measures of central a. Parametric tests

tendency (need to meet stat III. Statistical Notations for Single Scores
 e.g. Mean assumptions about
population): e.g. x = 12, 10, 8, 7, 5
b. Measures of  e.g. t-test, F-
variability test / ANOVA
 e.g. Std  Linear
deviation egression

c. Measures of relative b. Non-parametric tests

standing in a (no stat
group assumptions about
 e.g. Percentile population):
ranks  e.g. Chi-
square, Mann-
Whitney
 Non-linear
regression
IV. Statistical Notations for Pairs of Scores

Statistical Notations

A number of symbols are used in statistics, but there

is no absolute conformity, thus, statistical notations
vary.

I. Data Presentation

Single group data may be expressed either: Conventions for Calculations

a. Horizontally: X = 12,10,8,7,5 I. Rounding Numbers

b. Vertically: X To round to Nth place, refer to the number succeeding

12 (to the right of) N:
10
8
7
5

Multi-group data may be expressed either:

a. Spread (vertical):

Group 1 Group 2
Hours 10 5
of 8 13
Sleep 6 8
12 9

b. Stacked:

Hours of Sleep Group Number

10 1 II. Sequence of Operations
8 1
6 1 Apply PEMDAS in carrying out calculations:
12 1
5 2  P = operations in parentheses are calculated
13 2 first
8 2  E = followed by exponents
9 2  MDAS = followed by the usual MDAS

II. Statistical Notations

Rules of Summation
PSYCH STATS CHAPTER 1: INTRO TO PSYCH STATS