Title: PETROLEUM ENGINEERING 689 Special Topics in Unconventional Resource Reserves Lecture 1B Descriptive
1PETROLEUM ENGINEERING 689Special Topics
inUnconventional Resource ReservesLecture
1BDescriptive StatisticsTexas AM
University - Spring 2007
2Descriptive Statistics
- Part A
- Measures of Central Tendency
3Learning Objectives
- Determine the mean, median, and mode of a data
set - Determine geometric, harmonic, and quadratic
means of a data set - Determine weighted averages of data sets
- Determine the range, standard deviation,
variance, mean absolute deviation, and
coefficient of variation for a data set
4Definitions
- Population data set that contains all possible
items of interest - Sample data set that contains only a few random
or otherwise representative elements of a data
set - Measure of central tendency most likely value of
a data set - Parameters measures of central tendency and
other statistical characteristics that describe a
population - Statistics corresponding measures and
statistical characteristics that describe a sample
5Mean
- Also called arithmetic mean
- Symbol µ represents population mean
- Symbol represents sample mean
- Commonly called average
- Calculated by adding values of all items in data
set and dividing by total number of items in set
6Properties of Mean
- Sum of deviations from mean is zero
- Sum of squared deviations minimized when
deviations are measured from mean
Mean may be influenced by extreme values.
7Median
- Central value in array of a data set
- Odd number of elements ? actual data element in
middle of set - Even number of elements ? arithmetic average of
the two data elements in middle of array - Found by arranging elements of data set in
ascending or descending order and identifying
midpoint
8Properties of the Median
- Not influenced by extreme values
- For perfectly symmetrical data set, median equals
mean
9Mode
- Most frequently occurring data element in data
set - Poor measure of central tendency in most
casesdoes not take into account values of other
data elements
10Properties of the Mode
- Data set may have more than one mode e.g.,
bimodal (two modes) - Unaffected by extreme values
11Calculate Footage Drilled
- 20 bits drilled 2,013 ft
- Determine mean, median, mode
Mode (most frequent)
12Calculate Footage Drilled
- 20 bits drilled 2,013 ft
- Determine mean, median, mode
13Geometric Mean (Gm)
- Nth root of product of individual data elements
of data set with N elements
- Calculation simplified using logarithms
- In terms of natural logarithms
14Properties of the Geometric Mean
- Biased toward smaller values appropriate for
skewed data sets (asymmetrical distributions) - Not affected as much as arithmetic mean by
extreme values - Undefined for data sets with negative or zero
values
15Harmonic Mean, Hm
- Reciprocal of arithmetic mean of reciprocals of
data elements in data set
16Quadratic Mean, Qm (Root Mean Square)
17Weighted Average
- Averages in which data elements are weighted by
frequency of occurrence
18Weighted Average
- Weighted geometric mean (Gwm)
- Weighted harmonic mean (Hwm)
19Calculate Footage Drilled
- Determine geometric, harmonic, quadratic means of
bit record
20Calculate Footage Drilled
- Determine geometric, harmonic, quadratic means of
bit record
21Calculate Footage Drilled
- Determine geometric, harmonic, quadratic means of
bit record
22Calculate Footage Drilled
- Determine geometric, harmonic, quadratic means of
bit record
23Calculate Weighted-Average Porosity
- 22-ft pay zone
- Calculate weighted-average porosity
24Calculate Weighted-Average Cost of Capital
- Company will invest in 500,000 project
- 150,000 equity at cost of 8
- 350,000 long-term debt at 18
- Calculate weighted-average cost of capital
25Measures of Variability
- Range, R
- Difference between highest and lowest values in
data set
- Not particularly useful measure of dispersion,
since it uses only two values from data set
26Standard Deviation, s
- Measure of dispersion of data elements about mean
- Coupled with mean, provides more information
about data set than any other measure
27Variance, s 2
- Simply square of standard deviation
- Not used directly in descriptive statistics
28Mean Absolute Deviation, dm
- Average deviation of data from the mean over all
observations
- For symmetric (bell-shaped) distributions
29Coefficient of Variation, ?
- Ratio of standard deviation to mean of data set
- Expresses standard deviation as fraction of
percentage of mean
30Calculate Statistical Values for Drilling
- Determine range, standard deviation, for bit
record
2
31Calculate Statistical Values for Drilling
- Determine range, standard deviation, for bit
record
32Calculate Statistical Values for Drilling
- Determine range, standard deviation for bit record
2
33Calculate Statistical Values for Drilling
- Determine range, standard deviation for bit record
2
34Calculate Statistical Values for Drilling
- Determine range, standard deviation for bit record
2
35Calculate Statistical Values for Drilling
- Determine range, standard deviation for bit record
2
- Standard deviation (alternate method)
36Calculate Statistical Values for Drilling
- Determine variance, mean absolute deviation for
bit record
2
- Variance (square of standard deviation)
37Calculate Statistical Values for Drilling
- Determine variance, mean absolute deviation for
bit record
2
38Calculate Footage Drilled
- Determine coefficient of variation for bit record
2
39What Weve Learned
- Determine the mean, median, and mode of a data
set - Determine geometric, harmonic, and quadratic
means of a data set - Determine weighted averages of data sets
- Determine the range, standard deviation,
variance, mean absolute deviation, and
coefficient of variation for a data set
40Descriptive Statistics
41Descriptive Statistics
- Part B
- Working With Grouped Data
42Learning Objective
- Determine measures of central tendency and
variability for grouped data sets
43Grouping Data
- Condensing large data sets into groups simplifies
calculations of parameters - Steps in grouping
- Define classes for data set to be analyzed
- Determine frequency of data element appearances
in each class - Calculate absolute and relative frequency
- Calculate class mark (CM), or midpoint, for each
class - Proceed with calculation of parameters
44Guidelines for Defining Classes
- Number of classes should be between 5 and 20
- Define classes so that every element in data set
falls into one and only one class - No class should be empty
45Guidelines for Defining Classes
- Approximate number of classes (Nc) for data set
with N elements
- Class interval (CI) should be same for entire
data set
46Determining Mean
Frequency of individual class
Class mark (mid value of individual class)
47Determining Median, Mode
- Mode taken as CM of class with highest frequency
of data elements
48Determining Geometric, Harmonic Mean
49Determining Standard Deviation, Variance
50Calculate Parameters From Grouped Data
- Calculate measures of central tendency for
grouped drill-bit data
51Calculate Parameters From Grouped Data
- Calculate measures of central tendency for
grouped drill-bit data
52Calculate Parameters From Grouped Data
- Calculate measures of central tendency for
grouped drill-bit data
(compared to 100.65 ft)
53Calculate Parameters From Grouped Data
- Calculate measures of central tendency for
grouped drill-bit data
(compared to 103.5 ft)
54Calculate Parameters From Grouped Data
- Calculate measures of central tendency for
grouped drill-bit data
(compared to 97.97 ft)
55Calculate Parameters From Grouped Data
- Calculate measures of central tendency for
grouped drill-bit data
(compared to 97.97 ft)
56Calculate Parameters From Grouped Data
- Calculate measures of central tendency for
grouped drill-bit data
(compared to 22.74 ft)
57Calculate Parameters From Grouped Data
- Calculate measures of central tendency for
grouped drill-bit data
(compared to 517.19)
58What Weve Learned
- Determine measures of central tendency and
variability for grouped data sets
59Descriptive Statistics