Transcript Construction Engineering 221

Construction Engineering 221
Probability and statistics
Data description
• Random sample- every member of the
population has an equal chance of being
selected into the sample. Hard to do
(unlisted phone #’s, no permanent address)
• Can use random number generator
• Non-random is OK if biases are known and
reported
Data description
• Data from the sample represent an array
(rows and columns) in raw form- how it
might look on an Excel spreadsheet
• Can organize the data in various ways to
simple the array and make it more
“readable”
– Classification- like letter grades, use class
intervals to combine similar scores
Data description
– Class boundaries separate one class from
another
– Edges of boundaries are “class limits”, and
represent the data where most information is
“lost”
– Class size, interval, and boundary are arbitrary
– Class mark is the midpoint of the class limits
Data description
• Class frequencies are the number of scores within
each class
• The list of frequencies forms a distribution that
typically assumes one of several standard forms
(Normal, Chi Square, Poisson, Binomial)
• Cumulative distribution is a representation of
number of scores below. When divided by total,
the number represents a percentile rank
Data description
• Histogram is a common graphical
representation of a frequency distribution
• Frequency polygon connects the midpoints
of the histogram class limits and represents
data distribution as a line
• Ogive- cumulative distributions- normal
distribution will have an S-shaped ogive
Data description
• Common engineering distributions
Binomial
Power log
normal
Chi square
Poisson
Log normal
Normal
Weibull