MEC-6

Monitor & Control Processes

in the
Integration Management

1
Agenda
 Monitor & Control Project Work
 Perform Integrated Change Control

2
Monitor & Control Project Work - Inputs & Outputs

1. PMP 1. WPR
Monitor &
2. WPI 2. Change Requests
Control
3. OPA 3. PMP Updates
Project
4. Schedule Forecast 4. PD Updates
Work
5. Cost Forecast 5. OPA Updates
6. Validated Changes
7. EEF

3
Monitor & Control Project Work Inputs – 1/5
Input Details
Project • Communications management plan
Management Plan • Cost baseline
• Cost management plan
• Human resource management plan
• Process improvement plan
• Procurement management plan
• Scope baseline
• Quality management plan
• Requirements management plan
• Risk management plan
• Schedule baseline
• Schedule management plan
• Scope management plan
• Stakeholder management plan

4
Monitor & Control Project Work Inputs – 2/5
Input Details
Schedule Forecast • Derived from progress against the schedule baseline and
(from Control expressed in terms of schedule variance (SV) and
Schedule) schedule performance index (SPI)
• Forecasted Project Completion Date derived from Planned
Project Duration and SPI
• The forecast may be used to determine if the project is
still within defined tolerance ranges and identify any
necessary change requests.

Cost Forecast • Derived from progress against the cost baseline and
(from Control computed estimates to complete (ETC), expressed in
Cost) terms of cost variance (CV) and cost performance index
(CPI)
• Estimate at completion (EAC) can be compared to the
Budget at completion (BAC) to see if the project is still
within tolerance ranges or if a change request is required
• Forecast also ascertained through ETC and TCPI

5
Monitor & Control Project Work Inputs – 3/5
Input Details
Validated • Approved changes that result from the Perform Integrated Change
Changes Control process require Validation to ensure that the change was
appropriately implemented
• A validated change provides the necessary data to confirm that
the change was appropriately executed

WPI • WPI is the WPD collected from various controlling processes,

analyzed in context, and integrated based on relationships across
areas. Thus WPD data is transformed into WPI
• WPD in itself cannot be used in the decision-making process as it
has only out-of-context meaning. Work performance information,
however, is correlated and contextualized, and provides a sound
foundation for project decisions
• WPI is circulated through communication processes
• Examples: Status of deliverables, implementation status for
change requests, and forecasted estimates to complete

6
Monitor & Control Project Work Inputs – 4/5
Input Details
EEF • Governmental or industry standards (e.g., regulatory agency
regulations, codes of conduct, product standards, quality
standards, and workmanship standards)
• Organization work authorization systems
• Stakeholder risk tolerances
• Project Management Information System (e.g., an automated tool
suite, such as a scheduling software tool, a configuration
management system, an information collection and distribution
system, or web interfaces to other online automated systems).

7
Monitor & Control Project Work Inputs – 5/5
Input Details
OPA • Organisational communication requirements
• Financial controls procedures (e.g., time reporting, required
expenditure and disbursement reviews, accounting codes, and
standard contract provisions)
• Issue and defect management procedures defining issue and
defect controls, issue and defect identification, and resolution and
action item tracking
• Change control procedures, including those for scope, schedule,
cost, and quality variances
• Risk control procedures including risk categories, probability
definition and impact, and probability and impact matrix
• Process measurement database used to make available
measurement data on processes and products
• Lessons learned database

8
Monitor & Control Project Work Outputs … 1/3
Output Details
Change • Variances from comparing planned results to actual results
Requests • Change Requests expand, adjust, or reduce Project Scope, Product
Scope, or Qualty Requirements and Schedule or Cost Baselines
• Change Requests may necessitate the collection and documentation
of new requirements
• Change Requests can impact the project management plan, project
documents, or product deliverables
• Changes that meet the project’s change control criteria should go
through the integrated change control process established for the
project
• Changes may include:
- Corrective action—An intentional activity that realigns the
performance of the project work with the Proj Management Plan
- Preventive action—An intentional activity that ensures the future
performance of the project work is aligned with the Project
Management Plan
- Defect Repair—an intentional activity to modify a nonconforming
product or product component

9
Monitor & Control Project Work Outputs … 2/3
WPR • WPR are the physical or electronic representation of WPI compiled
in project documents, intended to generate decisions, actions, or
awareness
• Project information may be communicated verbally from person
to person. However, in order to record, store, and sometimes
distribute WPI, a physical or electronic representation in the form
of project documents is required
• WPR are a subset of project documents, which are intended to
create awareness and generate decisions or actions
• Specific work performance metrics may be defined at the start of
the project and included in the normal WPR provided to key
stakeholders
• Examples: status reports, memos, justifications, information
notes, recommendations, and updates

PMP/PD
Updates

10
Monitor & Control Proj Work – Tools & Techniques … 1/2
Tools & Techs Details
Expert To interpret the information provided by the monitor and
Judgement control processes, or to resolve any situation emerging from the
M&C process, expert judgement from professionals in the
industry may be sought

Project Provides access to automated tools, such as scheduling, cost,

Management and resourcing tools, performance indicators, databases, project
Information records, and financials used during the Monitor and Control
System Project Work process.

Meetings • May include project team members, stakeholders, and others

involved in or affected by the project
• Types: User groups, Review Meetings, Focal Groups etc

11
Monitor & Control Proj Work – Tools & Techniques … 2/2
Tools & Techs Details
Analytical To forecast potential outcomes based on possible
Techniques variations of project or environmental variables and their
relationships with other variables:
• Regression analysis
• Grouping methods
• Causal analysis
• Root cause analysis
• Forecasting methods (e.g., time series, scenario building,
simulation, etc.)
• Failure mode and effect analysis (FMEA)
• Fault tree analysis (FTA)
• Reserve analysis
• Trend analysis
• Earned value management (EVM)
• Variance analysis

12
M&C PW – Tools & Techniques - Grouping

• Grouping methods
• Causal analysis
• Root cause analysis Root Cause Analysis
• Failure Mode & Effect Analysis (FMEA)
• Fault Tree Analysis (FTA)

• Regression Analysis
• Forecasting methods
Trend Analysis, Smoothing,
• Trend analysis Forecasting
• Earned Value Management
• Variance Analysis

• Reserve analysis Risk Analysis

13
Grouping Methods

 Grouping methods are techniques for classifying observations into

meaningful categories
Example Defects on a deliverable, say a generator. Grouping by:
By basic engineering: Mechanical, Electrical, Chemical (battery) etc
By Deliverable area: Engine, Generator, Controls, Battery etc
By Load Conditions: Over, Heavy, Medium, Low, Under

14
Route Cause Analysis (RCA) & Causal Analysis

 A specific technique used to identify a problem, discover the

underlying causes that lead to it, and develop preventive action
 The root cause, once removed from the problem fault sequence,
prevents the final undesirable event from recurring
 Methods:
- Cause-and-effect diagrams/fishbone diagrams/Ishikawa
diagrams
- The 5 Ws (Whys)
- FMEA
- FTA
- Pareto etc

15
RCA – Cause & Effect Diagram
• Project: Improving the Quality of a University's MBA3.5 (3.5 yrs) Programme
• 1st Objective: Root cause the reason for low quality of MBA(3.5 yrs) graduates
Educational
Extent of Business
Background
Education
Weak FA/FSc
Weak BA/BSc/BCom
Roadmap
unextenable BA/BSc/BCom from
3.5 yrs vis-à-vis Low-Key Institutes
5.5/6.5 others BA/BSc in Low-Key subjects
Low Quality Product
in MBA (3.5 yrs)
Late Start Weak Inter-Personal Skills
Relatively
Early Finish Weak English
Time & Cost
Fixed

Other Reasons Personality

Years since FA/FSc/A-Lvl
0 1 2 3 4 5 6 7 8 9
BBA4/Masters MBA 2.5
BBA 4 MBA 1.5
BA/BSc/BCom MBA 3.5

16
RCA – Cause & Effect Diagram leading to Pareto Analysis
The possible root causes ascertained from the Cause & Effect diagram could next could be short-
listed through further RCA. Say a sample of 100 MBA(3.5yrs) students was surveyed which
returned the following data which could then be subjected to Pareto analysis to identify those
root-causes which would be responsible for 80% of the problem.
Raw Survey Data Pareto Tabulation
Root Causes Count Count %age Cum %age
1 Weak FA/FSc 90 97
5 21.1% 21.1%
2 Weak BA/BSc/BCom 18 95
6 20.7% 41.8%
3 BA/BSc/BCom from low-key institutes 7 90
1 19.6% 61.4%
4 BA/BSc/BCom in low-key subjects 15 88
8 19.2% 80.6%
5 Weak Inter-Personal Skills 97 18
2 3.9% 84.5%
6 Weak English 95 15
4 3.3% 87.8%
7 MBA3.5 roadmap unextendable 12 15
10 3.3% 91.1%
8 3.5 yrs vis-à-vis 5.5/6.5 others 88 12
7 2.6% 93.7%
9 Late start 12 12
9 2.6% 96.3%
10 Relatively early finish 15 10
11 2.2% 98.5%
11 Time & Cost fixed 10 7
3 1.5% 100.0%
17
RCA – Cause & Effect Diagram leading to Pareto Analysis
120 100%

90%

100
80%

70%
80

60%

60 50%

40%

40
30%

20%
20

10%

0 0%

18
RCA – the 5 Whys
• Problem: Excavator, an important source on the project, not starting

1st Why Why is the excavator not starting?

Flat battery
2nd Why Why Flat battery?
Alternator not working
3rd Why Why Alternator not working?
Drive belt from the engine broken
4th Why Why Drive belt broken?
Poor maintenance Part Not Available
5th Why Why Part Not Available?
Old excavator

19
Failure Mode & Effect Analysis (FMEA)

 Each potential failure Mode in every component of a product is

analyzed to determine its effect on the reliability of that
component and, by itself or in combination with other possible
failure modes, on the reliability and function of the deliverable,
product or system
 Primarily a Design/Planning Tool; also used for Monitoring &
Control where it provides a basis for identifying Root Failure
Causes and developing effective Corrective Actions

20
FMEA - Application
Monitoring & Control: Track changes to process-
incorporated to avoid potential failures

21
 22

FMEA
• Project: New Fuel-Efficient, Composite Material Commercial Plane (Boeing 787
Dreamliner)
• Deliverable: Lithium Ion Batteries

Product/Deliverable: Lithium Ion Battery FMEA Team: ABC (Chief Designer) Page No: 7 of 104
System: Electrical DEF (Head of Elect Dept) FMEA No. 1234
Sub-System: Back-Up Power GHI (Battery Specialist) Date: 7 Nov 2010
Part No: xxxx-xxxx-xxxx IJK (Configuration Manager)
Function Potential Potential Potential Current Recommend- RPN

Detection (D)
Failure Mode Effect(s) of Causes of Controls/ ed Actions

Occurrence
Severity (S)
Failure Failure Tests

(O)
Provide Overheating/ - Toxic Fumes 10 Short- 2 (Test No) 10 Redesign 200
back-up Meltdown in cabin circuiting in
power to - Explosion in Fuel Cells
aircraft battery
instrumen compt Recharging 3 (Test No) 10 Check Elect 300
tation - Fire Malfunction Charging Sys
Unbalanced 2 (Test No) 10 Redesign 200
Chemical
Reaction
Battery Compt 3 (Test No) 1 Check AC 30
AC failure system
FMEA
• Project: New Fuel-Efficient, Composite Material Commercial Plane
(Boeing 787 Dreamliner)
• Deliverable: Lithium Ion Batteries
• Problem: Over-heating & melt-down

 Battery Compartment

23
Fault Tree Analysis (FTA)

 A top-down, Deductive failure analysis in which an undesired state

of a system is analysed using Boolean logic to combine a series of
lower-level events
 Helps monitor and control Project Processes and Deliverables
Scope
 Functions as a diagnostic tool to identify and correct causes of
Process failing or a Deliverable not meeting the Scope either in
specifications or performance parameters

24
Fault Tree Analysis
• Project: New Commercial Plane
• Deliverable: Evacuation Chute
• Problem: Failure of Auto-Deployment

25
Fault Tree Analysis
Auto-
• Project: New Commercial Plane Deployment
• Deliverable: Evacuation Chute Failure

• Problem: Failure of Auto-Deployment

Power Signal Mech

Failure Failure Failure

Gen Faulty Missing Faulty

Fault Wiring Signal Signal
3

Part Lub Faulty Rx Servo Part Missing

Tx Fault
Failure Failure Wiring Fault Failure Failure Part
1 2 4 5 6 7 8 9

26
Quantitative Forecasting

Quantitative
Forecasting

Time Series Causal

Forecasting Forecasting

Moving Exponential Trend Auto

Average Smoothing Models Regression

27
Moving Average … 1/3
• Moving Average (Rolling Average or Running Average) is a
calculation to analyse data points by creating a series of averages
of different subsets of the full data set
• Variations include: Simple, Weighted, Centred, Exponential etc
• Moving Average is used to overcome irregular, random, seasonal
or cyclic variations
• Overcoming variations is called "smoothing“
• Moving Average is a smoothing process
• Smoothing by Moving Average is done by taking average of three
(or more) recent observations, then dropping the first observation
and advancing to the next one, and continuing the process till
getting to the period/unit for which forecast is required

28
Moving Average … 2/3
• Each new data point is included in the average as it becomes
available, and the oldest data point is discarded
• The number of observations averaged is referred to as the
“k” number; the constant number k is specified at the outset
• The smaller the number k, the more weight is given to recent
periods; the greater the number k, the less weight is given to
recent periods
• A large k is desirable when there are wide, infrequent
fluctuations in the series.
• A small k is most desirable when there are sudden shifts in
the level of series
• For quarterly data, a four-quarter moving average, MA(4),
eliminates or averages out seasonal effects
29
Moving Average … 3/3
• For monthly data, a 12-month moving average, MA(12),
eliminate or averages out seasonal effect
• Equal weights are assigned to each observation used in the
average

30
Forecasting Variations
Irregular
variation

Trend

Trend Long-term movement in data

Irregular variations Caused by unusual circumstances
Random variations Caused by chance

Cycle

Cycle Wave-like variations lasting more than one year

Seasonal Variations

Seasonality Short-term Regular variations in data 31

Simple Moving Average (SMA) – Understanding the Basic Concepts

Case A Case B
Simple and Moving Averages can be House Duration to House Duration to
used to forecast and smoothen data. # Complete # Complete
1 260 1 250
Example: On a multi-housing project, 2 245 2 280
the time of completion of the first 10
3 255 3
houses (H1 to H10) is indicated in the 230
tables, headed Case A and Case B. 4 246 4 220
What can be the forecasted duration 5 254 5 260
of House # 11 (H11) in each case?
6 243 6 250
7 253 7 260
8 242 8 230
9 254 9 220
10 248 10 240
11 ? 11 ?

32
SMA – Understanding the Basic Concepts
265 Case A
260
260 House Duration to
255
255
254 253 254 # Complete
250 1 260
245 248 2
245
246 245
240 243 242 3 255
235 4 246
230 5 254
225
6 243
220
H1 H2 H3 H4 H5 H6 H7 H8 H9 H10 H11 7 253
8 242
Forecasted Duration of H11 9 254
Mean 250.0 Fixed Average 10 248
Mean (minus 1st) 248.9 (the 1st House it took longer) 11 ?
Mean (last 3) 248.0 Moving Average at k = 3
Mean (last 4) 249.3 Moving Average at k = 4
Mean (last 5) 248.0 Moving Average at k = 5
33
SMA – Understanding the Basic Concepts
265261 Case B
260
260 f(x) = − 2.61818181818182
256 x + 264.4 House Duration to
255 253 # Complete
257
250 1 261
247
245 242 2 257
245 239
240 3
240 260
235 4 253
230 5 256
225
6 245
220
H1 H2 H3 H4 H5 H6 H7 H8 H9 H10 H11 7 247
Forecasted Duration of H11 8 240
Mean 9 242
250.0 Fixed Average
Mean (minus 1st three) 10 239
246.0 (first three houses took longer)
Mean (last 3) 11 ?
240.3 Moving Average at k = 3
Mean (last 4) 242.0 Moving Average at k = 4
Mean (last 5) 242.6 Moving Average at k = 5
Using Trend Line 235.6 (trend line considers all data) 34
SMA – How to work out Case A
House Duration to SMA SMA Error Squared Error Squared
# Complete (k=3) (k=4) (k=3) (k=4)
(a) (b) (c) (a-b)2 (a-c)2
1 260
2 245
3 255
4 246 253.3 53.8
5 254 248.7 251.5 28.4 6.3
6 243 251.7 250.0 75.1 49.0
7 253 247.7 249.5 28.4 12.3
8 242 250.0 249.0 64.0 49.0
9 254 246.0 248.0 64.0 36.0
10 248 249.7 248.0 2.8 0.0
11 248.0 249.3
Mean Square Error 6.7 5.0
(a-b)2/n (a-c)2/n
35
 265260
SMA – Graphical Representation 260
255
255 254 253 254
250
265 Case A 245
240 245 246
243 242
248

235
230
260
260 225
220
H1 H2 H3 H4 H5 H6 H7 H8 H9 H10 H11

255
255 254 254
253

250 249.25

248 248
245 246
245

243
242
240
Actual Durations SMA (k=3) SMA (k=4)

235
H1 H2 H3 H4 H5 H6 H7 H8 H9 H10 H11
36
Weighted Moving Average (WMA)
• WMA is used when it is required to give different weightage to
different data. For example it may be required to give more
weightage to recent data
• Example: In the original multi-housing project example (Case A), it
is required to forecast the duration of the 11th house by giving 1/2
weightage to the most recent house duration, 1/3 to the middle
duration and 1/6 to the earliest.

8 240
9 245
10 253
11 Forecast = 240/6 + 245x1/3 + 253x1/2 = 248.2

37
Centred Moving Average (CMA)
• CMA is used for a number of situations particularly when there is
a seasonal component, or when there is a requirement to use the
past data
• CMA can be computed, using data equally spaced on either side
of the point in the series where the mean is calculated
• When k is even, “smoothing of smoothing” is done

38
CMA Data as per Case A
k odd (3)
House Duration to SMA CMA CMA Error
# Complete (k=3) (k=3) Squared (k=3)
(a) (b) (c) (a-c)2
1 260
2 245 253.3 69.4
3 255 248.7 40.1
4 246 253.3 251.7 32.1
5 254 248.7 247.7 40.1
6 243 251.7 250.0 49.0
7 253 247.7 246.0 49.0
8 242 250.0 249.7 58.8
9 254 246.0 248.0 36.0
10 248 249.7
11 248.0
Mean Square Error (MSE) 6.6
(a-c)2/n
39
CMA Data as per Case A

265

260 Actual Durations SMA (k=3) CMA (k=3)

260

255
255 254 254
253

250
248.0

248
248
245 246
245

243
242
240

235
H1 H2 H3 H4 H5 H6 H7 H8 H9 H10 H11
40
CMA k even (4) Data as per Case A
House Dur to SMA (k=4) CMA (k=4) CMA Error
# Complete Sq (k=4)
(a) (c) (c) (a-c)2
1 260
2 245
2.5 251.5
3 255 250.8 26.3
3.5 250.0
4 246 249.8 15.0
4.5 249.5
5 254 251.5 249.3 15.0
5.5 249.0
6 243 250.0 248.5 54.4
6.5 248.0
7 253 249.5 248.0 19.1
7.5 248.0
8 242 249.0 248.6 33.1
8.5 249.3
9 254 248.0
10 248 248.0
11 249.3
Mean Square Error (MSE) 5.2
(a-c)2/n 41
CMA Data as per Case A

265

260 Actual Durations SMA (k=4) CMA (k=4)

260

255
255 254 254
253

250
248
249.25
248.625

245 246
245

243
242
240

235
H1 H2 H3 H4 H5 H6 H7 H8 H9 H10 H11
42
Exponential Moving Average (EMA)
• EMA weigh past observations using exponentially decreasing
weights as the observations get older; recent observations are
given relatively more weight than the older observations
• The amount of weight applied to the past observations, or the
degree of smoothing required, is determined by the “smoothing
constant”
• In summary, This is in contrast to the SMA, in which the same the
weights assigned to the observations are the same and are equal
to 1/n. In EMA, however, there are one or more smoothing
parameters to be determined (or estimated) and these choices
determine the weights assigned to the observations

43
EMA
• The exponential smoothing equation is:
Fn+1 =  yn + (1- )Fn
where Fn+1 = Forecast for the next unit (to be estimated)
a = Smoothing constant, such 0 <  ≤ 1
yn = Actual value of the most recent unit
Fn = Forecasted value of the most recent unit
• Expanding the Equation:
Fn+1 =  yn + (1- )Fn
[ yn-1 + (1- )Fn-1]
=  yn + (1- )
= yn + (1- )yn-1 + (1- )2Fn-1
= (1- )0yn + (1- )1yn-1 + (1- )2Fn-1
[ yn-2 + (1- )Fn-2]
= (1- )0yn + (1- )1yn-1 + (1- )2
= (1- )0yn + (1- )1yn-1 + (1- )2Yn-2+ (1- )3Fn-2
= (1- )0yn + (1- )1yn-1 + (1- )2Yn-2+ (1- )3Fn-2 ……… (1- )n-1y1
= [(1- )0yn + (1- )1yn-1 + (1- )2Yn-2+ (1- )3yn-3 ……… (1- )n-1y1]
44
EMA
• However, in application, EMA is a simple affair. All what is
required to be done is:
– Select a suitable smoothing constant ()
– Take the most recent observation (yn) and multiply it with the smoothing
constant
– Take what was the forecasted (Fn) value of the most recent observation/
event and multiply it with the complementary of the smoothing constant
i.e (1- )
– Add the two products; the sum is the forecasted value for the next unit
• If the forecasted value (Fn) of the recent most event is not
available, then:
– Start analysing the data from the start, or from where the last (Fn) is
available, by calculating Fn using the EMA equation
– Continue calculating Fn by applying the EMA equation until the forecasted
value of the target event is available
45
 46
EMA
# Actual Forecasted # Actual Forecasted
Observation Observation Observation Observation
(yn) (yn)
(Fn, ,=0.8) (Fn,=0.8)

1 260 1 260  260

260.0
2 245 260.0 2 245
248.0
3 255 248.0 3 255
253.6
246 253.6 246
247.5
254 247.5 254
252.7
243 252.7 243
244.9
253 244.9 253
251.4
n-2 242 251.4 n-2 242
243.9
n-1 254 243.9 n-1 254
252.0
n 248 252.0 n 248
248.8 248.8
n+1 n+1
EMA
• Small  (  0) would mean:
– Little consideration to actual/historical data, maximum consideration to
previously forecasted data
– Maximum smoothening of the data
• Large  (  1) would mean:
– Maximum consideration to actual/historical data, little consideration to
previously forecasted data
– Little smoothening of the data

47
 48
EMA – Example (original Case A)
• Consider the data for the original Case A
• yn & Fn for various values of  are tabulated:
H# Duration to Forecasted Duration (Fn)
Complete (yn)
=1  = 0.8  = 0.6  = 0.5  = 0.4  = 0.2  = 0.1  = 0.0
1
260  260.0 260.0 260.0 260.0 260.0 260.0 260.0 260.0
2 260.0 260.0 260.0 260.0 260.0 260.0 260.0 260.0
245
3 245.0 248.0 251.0 252.5 254.0 257.0 258.5 260.0
255
4 255.0 253.6 253.4 253.8 254.4 256.6 258.2 260.0
246
5 246.0 247.5 249.0 249.9 251.0 254.5 256.9 260.0
254
6 254.0 252.7 252.0 251.9 252.2 254.4 256.6 260.0
243
7 243.0 244.9 246.6 247.5 248.5 252.1 255.3 260.0
253
8 253.0 251.4 250.4 250.2 250.3 252.3 255.0 260.0
242
9 242.0 243.9 245.4 246.1 247.0 250.2 253.7 260.0
254
10 254.0 252.0 250.5 250.1 249.8 251.0 253.8 260.0
248
11 248.0 248.8 249.0 249.0 249.1 250.4 253.2 260.0
 49
EMA – Example (original Case A)
262

260 260.0

258

256

254
253.2
252
250.4
250
248.8
248 248.0
248
246

244 249.0
242
Yn Fn @ α=1.0 Fn @ α=0.8
240 Fn @ α=0.6 Fn @ α=0.5 Fn @ α=0.4
Fn @ α=0.2 Fn @ α=0.1 Fn @ α=0.0
238
H1 H2 H3 H4 H5 H6 H7 H8 H9 H10 H11
 50
EMA – Example (original Case B)
• Consider the data for the original Case B
• yn & Fn for various values of  are tabulated:
H# Duration to Forecasted Duration (Fn)
Complete (yn)
=1  = 0.8  = 0.6  = 0.5  = 0.4  = 0.2  = 0.1  = 0.0
1
261  261.0 261.0 261.0 261.0 261.0 261.0 261.0 261.0
2 261.0 261.0 261.0 261.0 261.0 261.0 261.0 261.0
257
3 257.0 257.8 258.6 259.0 259.4 260.2 260.6 261.0
260
4 260.0 259.6 259.4 259.5 259.6 260.2 260.5 261.0
253
5 253.0 254.3 255.6 256.3 257.0 258.7 259.8 261.0
256
6 256.0 255.7 255.8 256.1 256.6 258.2 259.4 261.0
245
7 245.0 247.1 249.3 250.6 252.0 255.5 258.0 261.0
247
8 247.0 247.0 247.9 248.8 250.0 253.8 256.9 261.0
240
9 240.0 241.4 243.2 244.4 246.0 251.1 255.2 261.0
242
10 242.0 241.9 242.5 243.2 244.4 249.3 253.9 261.0
239
11 239.0 239.6 240.4 241.1 242.2 247.2 252.4 261.0
 51
EMA – Example (original Case B)
262
261.0
260

258

256

254
252.4
252

250

248
Yn Fn @ α=1.0 247.2
246 Fn @ α=0.8 Fn @ α=0.6

244 Fn @ α=0.5 Fn @ α=0.4

Fn @ α=0.2 Fn @ α=0.1
242 242.2
Fn @ α=0.0 241.1
240.4
240 239.6
239 239.0
238
H1 H2 H3 H4 H5 H6 H7 H8 H9 H10 H11
Correlation & Regression
Correlation is a statistical method used to determine whether a
linear relationship between variables exists
Regression is a statistical method used to describe the nature of the
relationship between variables, that is, positive or negative, linear or
nonlinear

• Together, Correlation & Regression address these questions

statistically:
1. Are two or more variables linearly related?
2. If so, what is the strength of the relationship?
3. What type of relationship exists?
4. What kind of predictions can be made from the relationship?

52
 53
Equation of a Straight Line
14
y = a + bx 13
f(x) = 2 x + 3
where 12

11 y-diff 8
x = value of independent variable, Slope = x-diff = 4 = 2

y-differential (=8)
10
on the x-axis 9

y = value of dependent variable, 8

on the y-axis 7

6
a = intercept on the y-axis; fixed 5
cost, quantity etc 4 x-differential (=4)

b = slope of the line; ratio of 3

differential in y-values to 2
Intercept on Y-axis (= 3)
corresponding differential in 1

x-values -2 -1
0
0 1 2 3 4 5
-1

-2
 54
Correlation & Regression - Example
• The amount of cement consumed on a multi-housing project is a
function of the covered area of the house
• Independent Variable (x)  Covered Area (deca square meters)
Dependent Variable (y)  Cement consumed (deca bags)
• Data as follows:
x y

10 30
12 32
6 25
15 46
8 29
5 19

• Work out the Regression Line and the Correlation Coefficient (R)
 55
Correlation & Regression - Example

x y xy x2 y2

10 30 300 100 900

12 32 384 144 1,024
6 25 150 36 625
15 46 690 225 2,116
8 29 232 64 841
5 19 95 25 361
 56 181 1,851 594 5,867

R = 6x1,851 – 56x181 = 0.95

(6x594-562) (6x5,5867-1812)

= 181x594 – 56x 1,851 = 9.01

6x56-594
y = 2.27x +
= 6x1,851 – 56x181 = 2.27
6x594-562 9.014
Correlation & Regression - Example
Cement Bags (x10)
50

45
f(x) = 2.26635514018692 x + 9.01401869158879
40 R² = 0.900599953290171
Cement Bags (x10)

35

30

25

20

15
4 5 6 7 8 9 10 11 12 13 14 15 16
Covered Area (sq meter x 10 )
56
Correlation & Regression

57
Finding the Regression Line Equation & the “R”

58
 59
How to Work out “R” & Regression Equation
• Manually (like we did)
• Scientific Calculator
• Trend line on Chart
• Excel Sheet, manually with formula
• Excel Sheet, using Data Analysis Feature
• Softwares, eg Minitab
60