Unit 5: Data Analyses for

Decision Making
Chapter 01: Key Concepts of Data Analysis & Decision Making
Chapter 01: Key Concepts of Data Analysis & Decision
Making
1.1 Managing Data
1.2 Introduction to Data Analysis
1.3 Quantitative Data Analysis Basics
1.4 Qualitative Data Analysis Basics
1.5 Elements of Decision Analysis
Learning objectives
▪ Overview of Data Management, Data Analysis & Decision Making
▪ Basics of Quantitative Data Analysis
▪ Basics of Qualitative Data Analysis
▪ Elements of Decision Making
Managing Data

Types of data management

• Data processing (Key: Range of data source, integrate and organize data in
datasets)
• Data storage (Repository, Warehouse, data lakes)
• Data cleansing (Process of fixing incorrect, incomplete, duplicate or error in
dataset)
• Data governance (Set of standard and business process, data privacy)
• Data security (Strategy to protect data from unauthorized access, theft etc.)
Data Quality Assurance
• Relevance (Appropriate data, serve purpose)
• Accuracy (No or less outlier)
• Consistency of data (Internal and external validity)
• Timeliness (Up to date)
• Compliances (Legal obligation)

Four steps: Plan, Do, Check, and Act. we need: accurate and timely
information to manage services and accountability.
Introduction to Data Analysis

• Data analysis is guided by your performance management plan.

• It tells you what, when and how will you analyze data and how will
you present results.
• Computer package MS Excel or software like SPSS, STATA, R, Python
are used for data analysis.
• Qualitative analysis is most often done by reading through qualitative
data in the form of data transcripts, such as notes from focus group
discussions or interviews, to identify themes that emerge from the
data. This process is called content analysis or thematic analysis
Qualitative Data Analysis Software

• Qualitative Data Analysis Software provides tools that helps with

qualitative research such as transcription analysis, coding and text
interpretation, recursive abstraction, content analysis, and discourse
analysis.
• NVivo, ATLAS.ti, Provalis Research Text Analytics Software, Quirkos,
MAXQDA, Dedoose, Raven's Eye, Qiqqa, webQDA, HyperRESEARCH,
Transana, F4analyse, Annotations, Datagrav are some of the Top
Qualitative Data Analysis Software.
• Data collection and subsequent analysis, visualization and
interpretation activities require time and input from the wider non-
MEAL project team. It is particularly important to coordinate your
data analysis with the overall project implementation calendar.
Quantitative Data Analysis Basics

• Two kinds of quantitative analysis: descriptive and inferential

• Descriptive data analysis is the analysis of a data set that helps you
describe, show or summarize data in a meaningful way so that
patterns might emerge.
• Inferential data analysis enables you to use data from samples to
make statistical generalizations about the populations from which the
data were drawn.
Understanding Quantitative Data

• Before beginning quantitative analysis, you need to understand what

kind of data you are working with.
• The kind of quantitative data determine the kind of statistical analysis
needed. So, understanding your data begins with understanding
variables.
• Variable: Any characteristic, number or quantity that can be
measured or counted. There are two categories of variables,
independent and dependent:
Variable

• Independent variables are just what they sound like; variables that
stand alone and are not changed by the other variables you are
looking at. Age, religion and ethnic group are all examples of
independent variables.
• Dependent variables are categories that depend on other factors. For
example, a dependent variable might be distance walked to collect
water or the incidence of waterborne disease.
• Different types of variables are measured or counted differently.
Example: Time measured using minutes or second, knowledge as test
score.
Data

• Proper analysis of your data requires you to understand it in terms of

its “level of measurement.”
• Data is classified into four fundamental levels of measurement:
nominal data, ordinal data, interval data, and ratio data.
• Nominal data example: Name, Gender, religion, Place etc.
• Ordinal data example: Scales measuring levels of satisfaction.
• Interval data example: Temperature, time. (Any numbers)
• Ratio data example: Height, weight (expressed in numbers but
absolute 0)
Analysing Quantitative Data Using Descriptive Statistics

• Measures of frequency - Display the number of occurrences of a

particular value(s) in a data set (frequency tables, cross-tabulation
tables).
• Measures of central tendency - Calculate the center value of data sets
(mean, median, mode).
• Measures of variability - Determine the extent to which data points in
the data set diverge from the average, and from each other (range,
standard deviation).
Exercise- Measures of frequency, central tendency,
variability
• Frequency tables, cross-tabulation tables
• Mean, median, mode
• Range, standard deviation

Go to google sheet and do an exercise

Inferential Analysis

• Descriptive statistics describes data (for example, a chart or graph)

and inferential statistics allows you to make predictions (“inferences”)
from that data.
• With inferential statistics, you take data from samples and make
generalizations about a population.
Test

1. Compare the significance of differences between groups:

Determining whether the differences that exist between subgroups
are large enough to matter.
2. Examine the significance of differences between variables to
determine correlation and, potentially, causation: Determining
whether your activities contributed to the changes you are seeing.
 Analysis Description Example questions
method

t-test The t-test compares the average for one subgroup “Is the average distance walked to collect water at the
against the average for another subgroup. end of the project significantly different from the
It can also compare differences in averages at two average distance walked at the beginning of the
points in time for the same subgroup. project?”
If the result of the test is statistically significant, you
can potentially consider it as a project impact.
Analysis of The ANOVA test compares the average result of three or “Does the average distance walked to collect water
variance more groups to determine the differences between vary significantly between villages 1, 2, 3 and 4?”
them.
Chi-square The chi-square test works with frequencies or You expect that the creation of new water points will
test percentages in the form of a cross-tabulation table. improve access to water and thus meet consumption
It helps you see the relationship (if any) between the needs for both large and small households.
variables and to know whether your results are what
you expect to see. “Is there a significant difference between the
responses of small and large households to
questions about household consumption needs?”
“How significant is this difference?”
Qualitative Data Analysis Basics

• Qualitative analysis is working with words that combine to become

ideas, opinions and impressions. There are fewer rules, and
approaches vary. In general, the objective of qualitative analysis is to
identify key themes and findings, including among subgroups if you
have them, from all the notes you have collected from your
interviews and focus group discussions.
• Qualitative analysis begins with the raw data, which can take many
forms. You might have recordings of interviews. You might have notes
from focus group discussions. The raw data need to be organized so
that they are easy to review.
Qualitative Data Analysis Basics

Once you have organized your raw data, you need to complete the
following steps:
Step 1: Code data: Begin to identify themes
Step 2: Index data: Match concepts and relevant quotations to the
codes you have identified.
Step 3: Frame data: Organizes your data according to categories that
are useful to you. – Matrix or Framework approach
Elements of Decision Analysis
Although decision making under uncertainty occurs in a wide variety of
contexts, all problems have three common elements:
1. The set of decisions (or strategies) available to the decision maker,
2. The set of possible outcomes and the probabilities of these outcomes,
3. A value model that prescribes monetary values for the various decision–
outcome combinations. Once these elements are known, the decision
maker can find an optimal decision, depending on the optimality
criterion chosen.
Before moving on to realistic business problems, we discuss the basic
elements of any decision analysis for a very simple problem.
We assume that a decision maker must choose among three decisions,
labelled D1, D2, and D3. Each of these decisions has three possible
outcomes, labelled O1, O2, and O3.
 Payoff tables

Possible decision criteria

Elements Expected monetary value
of Decision
Sensitivity analysis
Analysis
Decision tress

Risk profiles
 Payoff Table for Simple Decision
Problem outcomes
01 02 03
D1 10 10 10
Decision
D2 -10 20 30
D3 -30 30 80

At the time the decision must be made, the decision maker does not know which outcome will occur.
However, once the decision is made, the outcome will eventually be revealed, and a corresponding
payoff will be received. This payoff might actually be a cost, in which case it is indicated as a negative
value. The listing of payoffs for all decision–outcome pairs is called the payoff table.
If the decision maker chooses decision D2 and outcome O3 then occurs, a payoff of $30 is received.

A Payoff Table lists the payoff for each decision–outcome pair. Positive values correspond to rewards
(or gains) and negative values correspond to costs (or losses).
 Possible Decision Criteria
What do we mean when we call a decision the “best”
decision? We will eventually settle on one particular criterion
for making decisions, but we first explore some possibilities.
• maximizes the worst payoff-is appropriate for a very conservative
Problem outcomes (or pessimistic) decision maker.
01 02 03
• maximizes the best payoff- is appropriate for a risk taker (or
D1 10 10 10 optimist).
Decision
D2 -10 20 30
D3 -30 30 80
 Expected monetary value
• The EMV approach assesses probabilities for each outcome of
Problem outcomes each decision and then calculates the expected payoff from
01 02 03 each decision based on these probabilities.
D1 10 10 10 • This expected payoff, or EMV, is a weighted average of the
Decision payoffs in any given row of the payoff table, weighted by the
D2 -10 20 30
probabilities of the outcomes.
D3 -30 30 80
• You calculate the EMV for each decision, then choose the
decision with the largest EMV.
EMV for D1: 10 (a sure thing) • let’s assume that a decision maker assesses the probabilities of
EMV for D2: -10(0.3) + 20(0.5) + 30(0.2) = 13 the three outcomes in Table 1.9 as 0.3, 0.5, and 0.2 if decision
EMV for D3: -30(0.5) + 30(0.2) + 80(0.3) = 15 D2 is made, and as 0.5, 0.2, 0.3 if decision D3 is made.
• Researchers have generally concluded that EMV makes sense,
even for one-shot deals, as long as the monetary values are not
too large.
Sensitivity Analysis Simple decision problem under
Some of the quantities in a
uncertainty
decision analysis, particularly
the probabilities, are often
intelligent guesses at best. Outcome
01 02 03 EVM
Therefore, it is important,
especially in real-world business Decision D1 10 10 10 10
problems, to accompany any D2 -10 20 30 13
decision analysis with a
sensitivity analysis. D3 -30 30 80 15

Here we systematically vary

inputs to the problem to see Probabilities
how (or if) the outputs—the D2 0.3 0.5 0.2
EMVs and the best decision—
change. D3 0.5 0.2 0.3
Spreadsheet model of a simple decision
problem
Decision Trees
• A graphical tool called a decision tree has
been developed to represent decision
problems.
• Decision trees can be used for any
decision problems, but they are
particularly useful for the more complex
types.
• They clearly show the sequence of events
(decisions and outcomes), as well as
probabilities and monetary values.
Decision Tree Conventions
• Decision trees are composed of nodes (circles, squares, and triangles) and branches (lines).

• The nodes represent points in time. A decision node (a square) represents a time when the
decision maker makes a decision. A probability node (a circle) represents a time when the result
of an uncertain outcome becomes known. An end node (a triangle) indicates that the problem is
completed—all decisions have been made, all uncertainty has been resolved, and all payoffs and
costs have been incurred.
• (When people draw decision trees by hand, they often omit the actual triangles, as we have done
in Figure in previous slide. However, we still refer to the right-hand tips of the branches as the end
nodes.)
• Time proceeds from left to right. This means that any branches leading into a node (from the left)
have already occurred. Any branches leading out of a node (to the right) have not yet occurred.
Decision Tree Conventions
• Branches leading out of a decision node represent the possible decisions; the decision maker can
choose the preferred branch. Branches leading out of probability nodes represent the possible
outcomes of uncertain events; the decision maker has no control over which of these will occur.
• Probabilities are listed on probability branches. These probabilities are conditional on the events
that have already been observed (those to the left).
• Also, the probabilities on branches leading out of any probability node must sum to 1.
• Monetary values are shown to the right of the end nodes. (As we discuss shortly, some monetary
values are also placed under the branches where they occur in time.)
• EMVs are calculated through a “folding-back” process, discussed next. They are shown above the
various nodes. It is then customary to mark the optimal decision branch (es) in some way. We
have marked ours with a small notch.
 The ultimate payoffs appear next to the end nodes, to the right
of the probability branches.

• The decision node comes first (to the

left) because the decision maker
must make a decision before
observing the uncertain outcome.
• The probability nodes then follow
the decision branches, and the
probabilities appear above their
branches.

We put a notch on the D3 branch to

indicate that this decision is optimal. The EMVs above the probability nodes are
for the various decisions
Risk Profiles 100
80
Risk profile
0.6

0.5
60
0.4
40
0.3
20
0.2
0
1 2 3 0.1
-20
-40 0

D3 P

In our small example each decision leads to three possible monetary payoffs with various probabilities.
In more complex problems, the number of outcomes could be larger, maybe considerably larger. It is then
useful to represent the probability distribution of the monetary values for any decision graphically.
Specifically, we show a “spike” chart, where the spikes are located at the possible monetary values, and the
heights of the spikes correspond to the probabilities.
In decision-making contexts, this type of chart is called a risk profile. By looking at the risk profile for a
particular decision, you can see the risks and rewards involved. By comparing risk profiles for different
decisions, you can gain more insight into their relative strengths and weaknesses.
Risk Profiles 100
80
Risk profile
0.6

0.5
60
0.4
40
0.3
20
0.2
0
1 2 3 0.1
-20
-40 0

D3 P

The risk profile for a decision is a “spike” chart that represents the probability distribution of monetary
outcomes for this decision.
The risk profile for decision D3 appears in above figure. It shows that a loss of $30 has probability 0.5, a
gain of $30 has probability 0.2, and a gain of $80 has probability 0.3. The risk profile for decision D2 is
similar, except that its spikes are above the values _10, 20, and 30, and the risk profile for decision D1 is a
single spike of height 1 over the value 10.
Note that the EMV for any decision is a summary
measure of the complete risk profile—it is the Risk profile
mean of the corresponding probability distribution. 100 0.6
Therefore, when you use the EMV criterion for 80 0.5

making decisions, you are not using all of the 60

0.4
40
information in the risk profiles; you are comparing 20
0.3

only their means. Nevertheless, risk profiles can be 0

0.2
1 2 3
useful as extra information for making decisions. -20 0.1

-40 0
For example, a manager who sees too much risk in D3 P
the risk profile of the EMV-maximizing decision
might choose to override this decision and instead
choose a somewhat less risky alternative.