Data Classification Assignment

Geog 3530: Cartography & GIS - Peterson
University of Nebraska at Omaha

Purpose: This assignment will cover the basic procedures used to classify data. Students will be asked to classify data on percent black population by census tract in Omaha in four different ways. Following this, one of the classification methods will be chosen and mapped.

Data: The following data is for Douglas county which has a Federal Information Processing System (FIPS) code of 31055. The FIPS code for Nebraska is 31. Census tracts within Douglas county are labelled from 2 to 74.68. There are about 139 census tracts for Omaha.

Excel file

 
TRACT NAME POP BL Percent BL
000200 Census Tract 2 4026 834 20.72
000300 Census Tract 3 2618 1742 66.54
000400 Census Tract 4 2386 101 4.23
000500 Census Tract 5 1652 440 26.63
000600 Census Tract 6 1551 1019 65.70
000700 Census Tract 7 1409 1237 87.79
000800 Census Tract 8 2011 1686 83.84
001100 Census Tract 11 2894 2350 81.20
001200 Census Tract 12 2643 2024 76.58
001600 Census Tract 16 2684 257 9.58
001800 Census Tract 18 3011 740 24.58
001900 Census Tract 19 1558 265 17.01
002000 Census Tract 20 3145 52 1.65
002100 Census Tract 21 2277 94 4.13
002200 Census Tract 22 1401 17 1.21
002300 Census Tract 23 2305 40 1.74
002400 Census Tract 24 3353 128 3.82
002500 Census Tract 25 2580 38 1.47
002600 Census Tract 26 2313 67 2.90
002700 Census Tract 27 2440 61 2.50
002800 Census Tract 28 3069 99 3.23
002900 Census Tract 29 5038 1153 22.89
003000 Census Tract 30 5998 160 2.67
003100 Census Tract 31 3139 102 3.25
003200 Census Tract 32 2403 86 3.58
003300 Census Tract 33 2210 70 3.17
003401 Census Tract 34.01 3425 131 3.82
003402 Census Tract 34.02 2533 38 1.50
003500 Census Tract 35 4326 87 2.01
003600 Census Tract 36 4432 55 1.24
003700 Census Tract 37 2542 26 1.02
003800 Census Tract 38 4489 308 6.86
003900 Census Tract 39 2942 290 9.86
004000 Census Tract 40 2994 309 10.32
004200 Census Tract 42 1556 100 6.43
004300 Census Tract 43 2928 175 5.98
004400 Census Tract 44 1565 32 2.04
004500 Census Tract 45 3069 28 0.91
004600 Census Tract 46 2419 33 1.36
004700 Census Tract 47 2788 34 1.22
004800 Census Tract 48 4423 344 7.78
004900 Census Tract 49 4627 875 18.91
005000 Census Tract 50 4130 548 13.27
005100 Census Tract 51 2853 1128 39.54
005200 Census Tract 52 1822 1493 81.94
005300 Census Tract 53 2158 1401 64.92
005400 Census Tract 54 3382 1696 50.15
005500 Census Tract 55 5211 251 4.82
005600 Census Tract 56 4166 215 5.16
005700 Census Tract 57 4445 592 13.32
005800 Census Tract 58 4863 1729 35.55
005901 Census Tract 59.01 2654 2017 76.00
005902 Census Tract 59.02 2228 1882 84.47
006000 Census Tract 60 4342 2924 67.34
006101 Census Tract 61.01 2553 1658 64.94
006102 Census Tract 61.02 4197 2272 54.13
006202 Census Tract 62.02 5166 1518 29.38
006301 Census Tract 63.01 2855 1636 57.30
006302 Census Tract 63.02 3968 2353 59.30
006303 Census Tract 63.03 2928 930 31.76
006400 Census Tract 64 5052 171 3.38
006503 Census Tract 65.03 2644 159 6.01
006504 Census Tract 65.04 3703 147 3.97
006505 Census Tract 65.05 2068 484 23.40
006506 Census Tract 65.06 3299 918 27.83
006602 Census Tract 66.02 5349 159 2.97
006603 Census Tract 66.03 2473 261 10.55
006604 Census Tract 66.04 3977 124 3.12
006701 Census Tract 67.01 3904 83 2.13
006703 Census Tract 67.03 3137 80 2.55
006704 Census Tract 67.04 1713 16 0.93
006803 Census Tract 68.03 2094 36 1.72
006804 Census Tract 68.04 1524 0 0.01
006805 Census Tract 68.05 3326 35 1.05
006806 Census Tract 68.06 2907 220 7.57
006903 Census Tract 69.03 2500 22 0.88
006904 Census Tract 69.04 3954 32 0.81
006905 Census Tract 69.05 1881 9 0.48
006906 Census Tract 69.06 3182 20 0.63
007001 Census Tract 70.01 3153 216 6.85
007002 Census Tract 70.02 3424 53 1.55
007003 Census Tract 70.03 2331 31 1.33
007101 Census Tract 71.01 3110 28 0.90
007102 Census Tract 71.02 3554 46 1.29
007303 Census Tract 73.03 2916 41 1.41
007304 Census Tract 73.04 1592 181 11.37
007307 Census Tract 73.07 3337 93 2.79
007308 Census Tract 73.08 1812 62 3.42
007309 Census Tract 73.09 2175 90 4.14
007310 Census Tract 73.10 2916 552 18.93
007311 Census Tract 73.11 2841 302 10.63
007312 Census Tract 73.12 1817 168 9.25
007313 Census Tract 73.13 3187 412 12.93
007405 Census Tract 74.05 2042 193 9.45
007406 Census Tract 74.06 5355 57 1.06
007407 Census Tract 74.07 3195 63 1.97
007408 Census Tract 74.08 4311 119 2.76
007409 Census Tract 74.09 2461 38 1.54
007424 Census Tract 74.24 2963 134 4.52
007429 Census Tract 74.29 3329 13 0.39
007430 Census Tract 74.30 3326 62 1.86
007431 Census Tract 74.31 3519 74 2.10
007432 Census Tract 74.32 2923 83 2.84
007433 Census Tract 74.33 4459 309 6.93
007434 Census Tract 74.34 3472 388 11.18
007435 Census Tract 74.35 3581 151 4.22
007436 Census Tract 74.36 4467 304 6.81
007437 Census Tract 74.37 5291 80 1.51
007438 Census Tract 74.38 1975 24 1.22
007439 Census Tract 74.39 4957 175 3.53
007440 Census Tract 74.40 1694 109 6.43
007441 Census Tract 74.41 3074 22 0.72
007442 Census Tract 74.42 5354 45 0.84
007443 Census Tract 74.43 3551 82 2.31
007444 Census Tract 74.44 4291 223 5.20
007445 Census Tract 74.45 2530 199 7.87
007446 Census Tract 74.46 4531 64 1.41
007447 Census Tract 74.47 3026 45 1.49
007448 Census Tract 74.48 2872 17 0.59
007449 Census Tract 74.49 2047 8 0.39
007450 Census Tract 74.50 3820 44 1.15
007451 Census Tract 74.51 4807 48 1.00
007452 Census Tract 74.52 3817 19 0.50
007453 Census Tract 74.53 3755 4 0.11
007454 Census Tract 74.54 4193 28 0.67
007455 Census Tract 74.55 1655 20 1.21
007456 Census Tract 74.56 2393 15 0.63
007457 Census Tract 74.57 2759 80 2.90
007458 Census Tract 74.58 3192 120 3.76
007459 Census Tract 74.59 2980 27 0.91
007460 Census Tract 74.60 2305 22 0.95
007461 Census Tract 74.61 3179 4 0.13
007462 Census Tract 74.62 5042 51 1.01
007463 Census Tract 74.63 4888 104 2.13
007464 Census Tract 74.64 2794 27 0.97
007465 Census Tract 74.65 3856 57 1.48
007466 Census Tract 74.66 6220 114 1.83
007467 Census Tract 74.67 5107 140 2.74
007468 Census Tract 74.68 2532 46 1.82

Map (click for PDF version)

Data Classification

Four methods of data classification are used here:

Equal Interval: This approach creates five equal steps in the data range. This equal data step is determined by dividing the difference between the maximum and minimum value by the number os classes. So, if the maximum is 100 and the minimum is 0, the equal step is 20. The class breaks would then be:

80.10 – 100.00
60.01 – 80.0 0
40.01 – 60.00
20.01 – 40.00
0.0 – 20.00

 

Quantile: The quantile method divides the distribution into an equal number of observations. For example, if there are 100 observations (counties, census tracts) and we want five classes (quintile), then we would have 20 observations in each class.

    Number of Observations
mid-point betweent 80th to 81st obs + 0.01 – maximum value
20
mid-point between 60th to 61st obs + 0.01 – mid-point between 80th to 81st obs
20
mid-point between 40th to 41st obs + 0.01 – mid-point between 60th to 61st obs
20
mid-point between 20th to 21st obs + 0.01 – mid-point between 40th to 41st obs
20
minimum value – mid-point between 20th to 21st obs
20

 

Standard Deviation: Classification based divisions of the standard deviation such that the area under the normal curve is divided into equal sections. These divisions are based on the chi-square values. For a five class map, the class breaks are 0.84 and 0.26 from the mean in both positive and negative directions.

    Number of Observations
(Mean + 0.84 * stdev) + 0.01 – maximum value
about 20% of values
(Mean + 0.26 * stdev) + 0.01 – (Mean + 0.84 * stdev)
about 20% of values
(Mean - 0.26 * stdev) + 0.01 – (Mean + 0.26 * stdev)
about 20% of values
(Mean - 0.84 * stdev) + 0.01 – (Mean - 0.26 * stdev)
about 20% of values
minimum value - (Mean - 0.84 * stdev)
about 20% of values

If the number of observations in each class is significantly greater or less than 20% of the overall number of observations (for five classes), then the data is, by definition, skewed. To convert the skewed data to a normal distribution, change each value to the Log10 equivalent and perform the standard deviation classificaton again. Convert the Log10 values back to normal using the 10 to the power of x formula, where x is the Log10 value.

 

Natural Break: Usually simply selected arbitrarily based on largest perceived "breaks" between values, Done systematically here by finding the mid-point between the four largest differences in the ranked data.

   
(mid-point between the two values that represent one of the 4 largest differences in the ranked values) + 0.01 – maximum value
(mid-point between the two values that represent one of the 4 largest differences in the ranked values) + 0.01 – (mid-point between the two values that represent one of the 4 largest differences in the ranked values)
(mid-point between the two values that represent one of the 4 largest differences in the ranked values) + 0.01 – (mid-point between the two values that represent one of the 4 largest differences in the ranked values)
(mid-point between the two values that represent one of the 4 largest differences in the ranked values) + 0.01 – (mid-point between the two values that represent one of the 4 largest differences in the ranked values)
minimum value – (mid-point between the two values that represent one of the 4 largest differences in the ranked values)

 

Assignment:

1) Copy the tract numbers and percent black values into Excel.

2) Complete all blanks in "classif.htm" and place in your folder.

3) Choose one of the classification methods and create a map in ArcGIS of the Omaha census tracts using the OmahaTracts.mdb (zipped) file.
Do not use the ArcGISs classification option. Instead, enter the class number for each census tract and assign the class breaks values that you calculated in ArcGIS legend editor.
Submit as OmahaClassif.pdf