PSYCHOLOGICAL STATISTICS

Introduction to Statistics for  Statistical inference - The process of making an

Psychology and Basic Concepts of Statistics estimate, prediction, or decision based upon sample
data
Data - Information collected to gain knowledge about a Types of Data
field or to answer a question of interest. 1. Qualitative - Categorical, i.e., data represents
categories
Data Sources include: e.g. – religion, gender, etc.
 Surveys (Mail, Telephone, Internet) 2. Quantitative - Data are numerical values
 Experiments (These days, huge data sets routinely  Discrete(countable) - Counts of things. e.g. -
generated in other ways in business, government, at Number of bedrooms
Internet sites)  Continuous(interval) - Measurements. e.g. -
Appraised value or square footage
Statistics- Science of collection, presentation, analysis, and
reasonable interpretation of data. (the art and science of Variable - any characteristic of an individual or entity. A
learning from data) variable can take different values for different individuals.
1. Nominal - Categorical (e.g., gender). The only
Statistics provides methods for: operation that can be applied to Nominal variables is
enumeration.
Design - Planning/Implementing a study 2. Ordinal - Variables with an inherent rank or order,
Sample survey or experiment? (e.g. mild, moderate, severe). Can be compared for
How to choose people (subjects) for the study, and how equality, or greater or less, but not how much greater
many? or less.
3. Interval - Values of the variable are ordered as in
Description – Graphical and numerical methods for Ordinal, and additionally, differences between values
summarizing the data are meaningful, however, the scale is not absolutely
anchored. (e.g. Calendar dates and temperatures)
Inference – Methods for making predictions about a 4. Ratio - Variables with all properties of Interval plus
population (total set of subjects of interest), based on a an absolute, non-arbitrary zero point, (e.g. age,
sample (subset of the sample on which study collects data) weight, temperature (Kelvin)).

Qualitative/non-parametric (you can’t use statistical

Population – total set
tools) – Nominal & Ordinal
of subject of interest.

Quantitative/parametric (you can use statistical tools) –

Interval & Ratio
Sample –
subset of the sample
Numerical Descriptive Measures: Notation
on which study
collects data

Normal Distribution –
equally/normally
distributed

Mean Descriptive statistics - statistical procedures used to

summarize, organize, and simplify data. describe
Basic Concepts
 Population - The set of all items of interest in a Inferential statistics - consist of techniques that allow us
statistical problem to study samples and then make generalizations about the
 Parameter - A descriptive measure of a population populations from which they were selected.
 Sample - A set of items drawn from a population
 Statistic - A descriptive measure of a sample
PSYCHOLOGICAL STATISTICS

Sampling error - the naturally occurring discrepancy, or then measuring the second variable to obtain scores
error, that exists between a sample statistic and the for each group
corresponding population parameter. The Experimental Method
 To demonstrate a cause-and-effect relationship
Do college students learn better by studying text on between two variables.
printed pages or on a computer screen?  An experiment attempts to show that changing the
value of one variable causes changes to occur in the
second variable.
 Characteristics:

Two general categories of variables that researchers must

consider:

Three basic techniques to control other variables:

1. Random assignment - each participant has an equal
chance of being assigned to each of the treatment
conditions
2. Matching - ensure equivalent groups or equivalent
environment
3. Holding them constant - holding age, environment,
etc. constant

One variable is manipulated while another variable is

observed and measured….
Data Structures, Research Methods, and Statistics
Independent variable is the variable that is manipulated
by the researcher.
Data structure I. Measuring two variables for each
individual: The correlational method
Dependent variable is the variable that is observed to
 two different variables are observed naturally to
assess the effect of the treatment.
determine whether there is a relationship between
them.
Sampling Techniques
Limitations of the correlational method
 The results from a correlational study can SAMPLING
demonstrate the existence of a relationship between  A sample is “a smaller (but hopefully representative)
two variables, but they do not provide an collection of units from a population used to
explanation for the relationship. determine truths about that population” (Field, 2005)
 a correlational study cannot demonstrate a cause-  Why sample?
and-effect relationship.  Resources (time, money) and workload
 To demonstrate a cause-and-effect relationship  Gives results with known accuracy that can be
between two variables, researchers must use the calculated mathematically
experimental method.  The sampling frame is the list from which the
potential respondents are drawn
Data structure II. Comparing two (or more) groups of  Registrar’s office
scores: Experimental and nonexperimental methods  Class rosters
 the relationship between variables is examined by  Must assess sampling frame errors
using one of the variables to define the groups, and What is your population of interest?
 To whom do you want to generalize your results?
PSYCHOLOGICAL STATISTICS

• All doctors select every kth (such as every 50th) element in the
• School children population.
• Indians
• Women aged 15-45 years
• Other
 Can you sample the entire population?
 3 factors that influence sample representative-ness
 Sampling procedure
 Sample size  Advantages:
 Participation (response)  Sample easy to select.
 •When might you sample the entire population?  Suitable sample can be identified easily.
 When your population is very small  Sample evenly spread over entire reference
 When you have extensive resources population.
 When you don’t expect a very high response
 Disadvantages:
SAMPLING BREAKDOWN  Sample may be biased if hidden periodicity in
population coincides with that of selection

3. Stratified
 subdivide the population into at least two
different subgroups (or strata) so that subjects
within the same subgroup share the same
characteristics (such as gender or age bracket),
then we draw a sample from each subgroup (or
stratum).

A probability sampling involves selecting members from a  Advantages:

population in such a way that each member of the  Using same sampling fraction for all strata
population has a known (but not necessarily the same) ensures proportionate representation in the
chance of being selected. sample
 Adequate representation of minority
1. Simple random subgroups of interest can be ensured.
 A simple random sample of n subjects is selected  Disadvantages:
in such a way that every possible sample of the  Sample of entire population has to be
same size n has the same chance of being prepared separately for each stratum.
chosen.  Can potentially require a larger sample than
 Applicable when population is small, would other methods.
homogeneous & readily available. 4. Cluster
 first divide the population area into sections (or
clusters), then randomly select some of those
clusters, and then choose all the members from those
selected clusters.

2. Systematic
 Systematic sampling, arranging the target population
according to some ordering scheme and then
selecting elements at regular intervals through that
ordered list. select some starting point and then
PSYCHOLOGICAL STATISTICS

 Advantages: Rules
 This can reduce travel and other administrative  Use multiple data collection methods
costs.  Use available data, but need to know
 Disadvantages:  how the measures were defined
 Sampling error is higher for a simple random  how the data were collected and cleaned
sample of same size.  the extent of missing data
 how accuracy of the data was ensured
Any sampling method where some elements of population  If must collect original data:
have no chance of selection is called non-probability  be sensitive to burden on others
sampling.  establish procedures and follow them (protocol)
 maintain accurate records of definitions and
Collection, Organization, and coding
Presentation of Data  verify accuracy of coding, data input

Data collection Structured Approach

 The process of gathering raw facts or evidence.  All data collected in the same way
 It can also be defined as gathering of information  Especially important for multi-site and cluster
(figures, words or responses) that describes some evaluations so you can compare
situation from which conclusions can be drawn.  Important when you need to make comparisons
with alternate interventions.
Data collection methods
Use Structured Approach When:
 Data-collection methods allow us to systematically
 need to address extent questions
collect information.
 have a large sample or population
 Data collection tools are instruments that are used
 know what needs to be measured
hand in hand with the data collection methods to
 need to show results numerically
collect/retrieve data.
 need to make comparisons across different sites or
interventions
Data collection techniques and tools
Semi-structured Approach
 Systematic and follow general procedures but data
are not collected in exactly the same way every time
 More open and fluid
 Does not follow a rigid script
 may ask for more detail
 people can tell what they
want in their own way

Types of data Use Semi-structured Approach when:

1. Secondary data: Data that previously may have  conducting exploratory work
been collected for other purposes but can be used in  seeking understanding, themes, and/or issues
the immediate study.  need narratives or stories
2. Primary data: Raw data that has been originally  want in-depth, rich, “backstage” information
collected by the researcher specifically for the study  seek to understand results of data that are
at hand unexpected
In conclusion, primary sources provide first hand data and
secondary sources provide second hand data. Characteristics of Good Measures
1. Relevance
Data Collection Strategies  Is the measure relevant?
No one best way: decision depends on:  Measure what is needed.
 What you need to know: numbers or stories 2. Credibility
 Where the data reside: environment, files, people  Is the measure credible and believable?
 Resources and time available 3. Internal Validity
 Complexity of the data to be collected  Is the measure valid?
 Frequency of data collection 4. Reliability
 Intended forms of data analysis  Is the measure reliable?
PSYCHOLOGICAL STATISTICS

 A measure’s precision and stability- extent to Data Collection Tools

which the same result would be obtained with 1. Participatory Methods - Involve groups or
repeated trials. communities heavily in data collection
Examples:
Quantitative Approach A. community meetings
 Data in numerical form  One of the most common participatory
 Data that can be precisely measured (age, cost, methods, must be well organized
length, height, area, volume, weight, speed, time, B. mapping
and temperature)  Drawing or using existing maps.
 Easier to analyze  Useful tool to involve stakeholders
a. increases understanding of the community
Qualitative Approach b. generates discussions, verifies secondary
 Data that deal with description sources of information, perceived changes
 Data that can be observed or self-reported, but not c. Types of mapping:
always precisely measured  natural resources, social, health, individual
 Less structured, easier to develop or civic assets, wealth, land use,
 Can provide “rich data” — detailed and widely demographics
applicable C. transect walks
 Is challenging to analyze  Evaluator walks around community
 Is labor intensive to collect observing people, surroundings, and
 Usually generates longer reports resources
 Need good observation skills
 Walk a transect line through a map of a
community — line should go through all
zones of the community
2. Records and Secondary Data
3. Observation - See what is happening
Degree of Structure of Observations
1. Structured: determine, before the observation,
precisely what will be observed before the
observation
2. Unstructured: select the method depending
upon the situation with no pre- conceived ideas
Obtrusive - data collection methods that directly obtain or a plan on what to observe
information from those being evaluated. e.g. interviews, 3. Semi-structured: a general idea of what to
surveys, focus groups observe but no specific plan
4. Surveys and Interviews
Unobtrusive data - collection methods that do not collect  Excellent for asking people about: perceptions,
information directly from those being evaluated. e.g., opinions, ideas
document analysis, Google Earth, observation at a  Less accurate for measuring behavior
distance, trash of the stars  Sample should be representative of the whole
 Big problem with response rates
How to Decide on Data Collection Approach  Structures for Surveys
 Choice depends on the situation a. Structured:
 Each technique is more appropriate in some  Precisely worded with a range of pre-
situations than others determined responses that the respondent
 Caution: All techniques are subject to bias can select
 Everyone asked exactly the same questions
Triangulation to Increase Accuracy of Data in exactly the same way, given exactly the
1. Triangulation of methods - collection of same same choices
information using different methods b. Semi-structured
2. Triangulation of sources - collection of same  Asks same general set of questions but
information from a variety of sources answers to the questions are predominantly
3. Triangulation of evaluators - collection of same open-ended
information from more than one evaluator 5. Focus Groups
 Type of qualitative research where small
homogenous groups of people are brought
PSYCHOLOGICAL STATISTICS

together to informally discuss specific topics

under the guidance of a moderator
 Purpose: to identify issues and themes, not just
interesting information, and not “counts”
6. Diaries, Journals, Self-reported Checklists
 Use when you want to capture information
about events in people’s daily lives
 Participants capture experiences in real- time not
later in a questionnaire
 Used to supplement other data collection
7. Expert Judgment
8. Other Tools

Data Collection Summary

 Choose more than one data collection technique
 No “best” tool
 Do not let the tool drive your work but rather choose
the right tool to address the evaluation question

Organization

Frequency Distribution - A frequency distribution is an

organized tabulation of the number of individuals located
in each category on the scale of measurement.
1. Simple Frequency Distribution - The most common
way to organize scores.
2. Grouped Frequency Distributions - When a set of
data covers a wide range of values, it is
unreasonable to list all of the individual scores in a
frequency distribution table. Hence, present groups
of scores rather than individual values.
Guidelines:
Step 1: number of class
Step 2: Class width = range/ no. of classes, range:
highest score-lowest score