PLAN MAKING & IMPLIMENTATION Flashcards
Population concepts
Migration - People moving and out of a community
Natural Increase - More births than deaths
Step Down Method - Use projections from larger entity
Cohort Survival Method - Multiple factors (cohorts)
Projecting populations
Electric meters / Utilities
Mathematical / Graphic techniques
Linear -
Exponential - Percentage
Gompertz Curve - Period of rapid growth then leveling
Birth Rate
2.1 birth rate necessary to maintain population size
Europe under 2, the US at 2.
Data Sources
US Bureau of census, Fedstats, National center for health statistics, state level dept of health
Employment Data
US Bureau of census, counter business patterns, census transport planning package
Economic Base Model - Multiplier Effect
How many additional jobs will be created by a new business? Is GM creating the new jobs? Not entirely, GM will cause other companies to create jobs.
Total Employment / Basic Employment
So every basic employee generates xxx employees. That person PLUS xxx employees.
Economic Base Model - Economic Base Analysis
Tries to determine the multiplier effect using location Quotient.
(Regional Employment in Industry / Total Regional Employment) VERSUS (National Employment in Industry / Total National Employment)
If location quotient > 1 then exporting jobs. If location quotient < 1 then importing jobs. Only basic activities can be imported (manufacturing). Non basic can’t be imported (hairdresser).
Shift-Share Analysis
Analyzes change in employment in given area / industry. Looks at two periods (ie between 1990 and 2000 a specific industry grew xxx percent while nationally xxx percent)
Input-Output Analysis
US bureau of economic analysis creates multipliers for different sectors. Certain jobs will have certain impacts.
Regional input-output modeling systems (RIMS). Provides multipliers by region. (ie casino employment in Las Vegas, how much additional jobs, population, etc.)
Statistics
Descriptive - No tests, just raw data.
Inferential - Use of sample information to tell us what is happening in the greater population and to measure the degree of uncertainty involved.
A sample statistic (english letters) to estimate or make an inference of a greater population parameter (greek letters).
Statistics - Speed of vehicles example
Descriptive - Speed of each vehicle collected
Inferential - When broken down into speed classes, we can see how many cars were travelling between xxx and xxx mph (frequency), we can see what portion or percent of total cars were travelling under xxx mph (cumulative frequency), we can see what portion of the traffic was travelling at xxx mph (relative frequency), we can see what total portion of traffic was travelling under xxx mph (cumulative relative frequency).
Categorized Data
Frequency - count of occurrences (cases) in each category
Cumulative Frequency - count of cases up to and including this category
Relative Frequency - each category’s proportion of total Sums to 1.0
Cumulative Relative Frequency - proportion of total
Histogram \ Bar Chart
Graphing of frequencies
Vertical Axis is frequency, relative frequency or cumulative frequency
Horizontal Axis is catogories
Bar chart of relative frequency, sum = 1.0
Statistical Vocabulary
Cases, entities, observations = what you are studying (ie cars, drivers)
Variables = Characteristics of the cases that are different for different cases (they “vary” ie type of car, speed, volume)
Discrete Variable = Take only some values in range (ie number of children in household or number of fire stations)
Continuous Variable = Variable that can take any value within the range (ie 5 miles or 5.342343 miles)
Measurement levels of Variables
Nominal - names categories, no value or order (ie race or marital status or car color)
Ordinal - ordered but don’t know size of gap between values (ie strongly agree to strongly disagree, 1st 2nd 3rd)
Interval - know order and size of gap (ie temperature)
Ratio - also has a true zero (ie income)
Measurement levels and Analysis
Levels matter because each level has a specific kind of analysis that make sens for it. (ie race is nominal variable African American =1, White =2, Asian = 3, etc)
Measures of Central Tendency
Central Tendency measures the center
Mean = The average
Median = The middle number of ordered list (best for skewed data ie income or speed where an outlier is throwing off the mean)
Mode = The most common (best for nominal or categorical data, okay for ordinal data, bad for continuous data)
Measures of Dispersion
Dispersion measures how spread out things are around that center
Range = maximum value - minimum value
Probability Concepts
Statistical Inference is based on being able to calculate probability of making an error
Probability of any event = proportion of that event out of all possibilities (ie relative frequency)
Relative Frequency (histogram) of a variable = probability distribution
Confidence Interval = 95% confident that the true mean of the population of car speeds fall between xxx and xxx mph. Can work backward to estimate needed sample size to achieve level of confidence required. Z table will tell you what sample size necessary.
Inference - Hypothesis Testing (7 steps)
- Try to reject the null - can never prove something true, but can prove something false.
- If you reject the null, then alternative must be true.
- If you cannot reject the null, then all you can say is that the null may be true.
Regression Analysis
Ordinary Least Squares Regression (OLS) - Linear Relationship.
OLS Example…
(dependent = speed; independent = # of cars)
Expectation: more cars means lower speed. Regression equation found: speed = 68 mph - 1.02 ( # of cars). 0 cars, speed = 68 mph. Lose 1.02 mph with every additional car. 67 cars would reduce speed to 0.
Speed as a function of the number of cars.
Used to show relationship between Dependent Variable and Independent Variable
Does not prove causality, only statistical relationship.
Sample Questions
OLS Example…
Is the housing stock substandard and to the degree it is substandard. The standard regression finds you have R squared of 0.43. What does this mean? 43% of the variation in the housing conditions is explained by the variables in analysis. NOT that 57% or 43% of housing stock is substandard.
Survey of household Example…
A survey of households to determine their attitudes about development had poor results. An accusation is made about inaccurate samples which don’t represent the city as a whole. What method would not improve the probability that a sample is accurate? Decrease sample size. By replicating the study, improving sampling techniques, using a less diverse population, and using multiple sample techniques would improve accuracy.
You hypothesize that introducing traffic calming measures will reduce average speed. You test your hypothesis, based on results what can you conclude regarding the null?
