Interpreting bivariate Local Moran statistics

The four quadrants in the Moran scatter plot provide a classification of four types of spatial autocorrelation. Areas that are significant are labelled with these categories in the "High-High/Low-Low" dataset produced in the Moran analysis, and are colored in the scatter plot to indicate their category (non significant locations are colored gray). The interpretation will be different for the univariate and bivariate Local Moran.

category	scatter plot quadrant and color	autocorrelation	interpretation
high-high	upper right red	positive	cluster - "I'm high in [ego variable at ego time] and my neighbors are high in [neighbor variable at neighbor time]."
high-low	lower right pink	negative	outlier - "I'm a high outlier in [ego variable at ego time] among low neighbors in [neighbor variable at neighbor time]."
low-low	lower left med. blue	positive	cluster - "I'm low in [ego variable at ego time] and my neighbors are low in [neighbor variable at neighbor time]."
low-high	upper left light blue	negative	outlier - "I'm a low outlier in [ego variable at ego time] among high neighbors in [neighbor variable at neighbor time]."

As stated above, in the Local Moran maps, only those regions with p-values below 0.05 are colored following the high/low coloring scheme described above (see map legend as well). Places with non-significant Local Moran statistics are colored gray, and if there were islands in the dataset they would be shown as missing values because they have no adjacent neighbors.