RESM 540

When are results statistically significant?

0.5 level

To determine if the explanatory variables are helping your model....

??? Determine if they are statistically significant or not Determine if the sign is appropriate Both A and B Only A

spatial autocorrelation

A measure of the degree to which a set of spatial features and their associated data values tend to be clustered together in space (positive spatial autocorrelation) or dispersed (negative spatial autocorrelation). *Closer things are more similar*

Identifies the mean direction, length, and geographic center for a set of lines

A+B: Creates standard deviational ellipses to summarize the spatial characteristics of geographic features Calculates central tendency, dispersion, and directional trends

The central feature command:

ALL: Needs data to be projected to be most accurate Selects a feature that is part of the mapped data Identifies the most centrally located feature Can be used on point, line, or polygon classes features.

The Measuring Geographic Distributions toolset addresses questions such as:

ALL: Where's the center? What's the shape and orientation of the data? How dispersed are the features?

Which of the following statements are true assuming the p-value returned using Gettis-Ord stat is small and statistically significant?

ALL: -The higher (or lower) the z-score the stronger the intensity of the clustering -A z-score near zero indicates no apparent clustering within the study area -A positive z-score indicates clustering of high values -A negative z-score indicates clustering of low values

Spatial statistics provides the ability to:

ALL:Rely on statistics to provide a more objective map display Highlight patterns and relationships in the data Use space or location as a fundamental component of the stats

Which of the following are true?

All: -the k function includes all distances to neighbors within a given distance -is applicable for emergency calls of bird nest studies -the k function is good if you are interested in how pattern changes at different scales of analysis

How does spatial statistics integrate space and spatial relationships directly into their mathematics?

By requiring the user to select a value for the Conceptualization of Spatial Relationships parameter prior to analysis

Average Nearest Neighbor

Calculates a nearest neighbor index based on the average distance from each feature to its nearest neighboring feature

The first step in a regression analysis is to:

Choose the variable you want to understand, predict, or model

Subjectivity can still exist in creating density maps in all but the following category:

Choosing the cell size

A biased model will have

Clustered residuals

If the index (ANN ratio) is less than 1 the pattern exhibits..

Clustering

Geographical data...

Contradicts statistical data

The Collect Events command does the following:

Counts the number of points at a location to allow a field "ICOUNT" to be used in the incremental spatial autocorrelation command

The main goal in regression analysis is to:

Develop a properly specified model

When running GWR, the GWR variables are the same as OLS, except

Do not include variables with little value variation

T/F: A cold spot is a location with a high values surrounding it.

False

T/F: A high positive z score for a feature indicates a cluster of similarly low values.

False

T/F: A hot spot is a location with low values surrounding it.

False

T/F: A low positive z score for a feature indicates a cluster of similarly high values

False

T/F: It is fine to have polygons that are used for the Ripley's K function to be in widely differing sizes

False

T/F: The OLS tool generates outputs that include a map of residuals and a summary report.

False

T/F: The most effective way to find the correct radius to use in the Point Density GIS command is by trial and error.

False

T/F: The regression residual map only shows under predictions from model results.

False

T/F: You can trust the adjusted R2 even if the other 5 tests have not been passed yet

False

Similarity search can be used to:

Find points that are most similar or most dissimilar based on feature attributes

All of the following are possible applications of using the similarity search tool except:

Finding other stores that are similar to a high producing one Finding other nest sites that were not productive THIS ONE: Finding water quality data that also has violations Locating the most suitable place to consider solar panels

Why do we need to have our data in a projected format?

Geographic coordinates do not accurately represent shape, area, distance and direction TRUE -Spatial statistics rely on distance relationships and calculations for testing the hypothesis of a random pattern TRUE -Projected coordinates systems minimize distortions TRUE

The linear directional mean command:

Identifies the mean direction, length, and geographic center for a set of lines

Average Nearest Neighbor

If: <1 = clustered >1 = dispersed calc: observed/expected

The standard distance command:

Measures the degree to which features are concentrated or dispersed around the geometric mean center

A regression model is inappropriately used when

NOT A (maybe (B, noa): You accept a lower R2 than appropriate You try to use a miss-specified model (it is missing explanatory variables) You use OLS instead of GWR None of the above

When we model spatial data, the two properties that make it very difficult to meet the assumptions and requirements of traditional statistical methods like OLS are:

Only B and C Geographic features that are spatially autocorrelated Nonstationarity of spatial data or regional variation

Which of the features below are two dimensional?

Polygon Grid cells

Whenever distance is a component in your analysis use

Projected!

In our ordinary least squares models, we can deal with spatial effects by

Resampling and spatial filtering

One way to test for spatial effects in your model is to

Run the Global Moran's I test for spatial autocorrelation of your model standardized residuals

T/F: For the average nearest neighbor statistic, the null hypothesis states that features are randomly distributed.

T

To test if explanatory variables are redundant (multicollinearity) use

The Variance Inflation Factor test

If you have large population value and other smaller values as attributes:

The population values need to be transformed so that population does not dominate the other attributes The transformation will automatically be performed within the tool B and C above

Why are so many density maps not done properly?

They do not include a statistically selected radius using spatial autocorrelation

T/F: A hotspot map of residuals can be used to help identify broad regional patterns and possible variables to include in the model.

True

T/F: A model is considered miss-specified due to spatial effects when the standardized residuals are spatially autocorrelated.

True

T/F: A reason to use the Ripleys K function is to find the distance in which features are statistically significant dispersed or clustered

True

T/F: Average nearest neighbor is best used when the features of interest have no interaction with each other.

True

T/F: Exploratory regression tool evaluates all possible combinations of the input candidate explanatory variables, looking for OLS models that best explain the dependent variable within the context of user-specified criteria.

True

T/F: Global statistics provide one value for the entire study area.

True

T/F: Identifying spatial clusters can help find the causes of the clustering.

True

T/F: If theory indicates that a variable is very important, it should be retained in the model even if it has an incorrect sign and/or is not statistically significant.

True

T/F: Local statistics provide probability and z scores for each feature location.

True

T/F: Null hypothesis states features are randomly distributed/no pattern or relationship

True

T/F: OLS is a global regression model and GWR is a local regression model

True

T/F: OLS is considered a global modeling approach and GWR a local modeling approach.

True

T/F: Spatial clusters can provide insights into variables to include in a model.

True

T/F: The Similarity Search command can find the most similar and the most dissimilar sites. For example, if you choose to report ten results of both most and least similar, the output will show the top ten similar sites (most similar = 1 and least similar = 10) and the top ten dissimilar sites (most dissimilar = -1 and least dissimilar = -10).

True

T/F: The average nearest neighbor statistic was limited for analyzing patterns in that it only considered the points themselves and no attributes of the points.

True

T/F: The input features and candidate features are two inputs in the Similarity Search command that can be created as layer files.

True

T/F: The z-scores and p-values are measures of statistical significance which tell you whether or not to reject the null hypothesis, feature by feature.

True

T/F: The z-scores and p-values indicate whether the apparent similarity (a spatial clustering of either high or low values) or dissimilarity (a spatial outlier) is more pronounced than one would expect in a random distribution.

True

T/F: We should always start with an OLS model because of the available diagnostic tests that are available

True

t/F: The incremental spatial autocorrelation tool requires an input attribute field.

True

Traditional Statistics are...

Void of influences Assume randomness and independence

When is the average nearest neighbor most effective?

When comparing different features in a fixed study area

ANN best used when...

features have interaction with each other

The best way to choose an appropriate distance band is to let the data show you with the incremental spatial autocorrelation tool

the first peak value should be chosen

The boundary correction method is needed because:

the number of neighbors on the edges of a study area can underestimated

To be sure we have clustered or dispersed pattern

we need to make sure out k function is outside of the confidence interval