MODULE 3A. Create and Interpret Tables

Transcription

1 MODULE 3A Create and Interpret Tables

2 Our Agenda Introductions, Curriculum Overview 5 min Before Analyzing the Data 20 min Basic Calculations 20 min Group Work 5 min Charts 20 min Group Work 20 min Maps 20 min

3 Introductions Let s share our name, organization, and experience with SAVI.

4 Where We Are in the Training Curriculum We are here, learning how interpret data means using excel tables.

5 What We Will Learn To use Excel to find what is important and meaningful in the data To use pivot tables to cross-tabulate, filter, summarize, and more To describe and compare your data

6 Before Analyzing the Data

7 Before Analyzing the Data We saved a working copy of the file, so we don t screw up the raw data. Work on three copies: Raw data: should be kept intact! Editable file: Copy of the raw data where all the possible analysis takes place. Final file: Clean and well formatted copy of the editable file (to be used for presentations, reports, etc.

8 Before Analyzing the Data Revisit your goals and objectives: What are you trying to accomplish? Do you need to modify your questions? Stay focused! It is easy to get lost in a sea of data and analysis.

9 Before Analyzing the Data Our questions: 1. How long have households been in their current house? Which neighborhoods are most transient? 2. Which neighborhood has the highest number of rental households? Percent rental households?

10 Before Analyzing the Data We looked at the metadata to see if we can use it to accomplish our goals.

11 Before Analyzing the Data We deleted the margin of error columns. They inform your decision making but clutter the analysis of the data.

12 Before Analyzing the Data We renamed columns in our working data to make it easier to use.

13 Basic Calculations

14 Basic Calculations Totals Middle tendency Ranges Percentages Frequency tables Conditions

15 Basic Calculations Totals It helps seeing the overall trends, used for other calculations Middle tendency Ranges Percentages Frequency tables Conditions

16 Basic Calculations Totals =SUM(cell 1: cell N) Middle tendency Ranges Percentages Frequency tables Conditions

17 Basic Calculations Totals Middle tendency Ways to see what are the data tendencies as well as how is the data is distributed Ranges Percentages Frequency tables Conditions

18 Basic Calculations Totals Middle tendency =Average(cell 1: cell N) =Median(cell 1: cell N) =Mode(cell 1: cell N) Ranges Percentages Frequency tables Conditions

19 Basic Calculations Totals Middle tendency Ranges Important to see what the full spectrum looks like. Is it pretty homogeneous or really diverse? Percentages Frequency tables Conditions

20 Basic Calculations Totals Middle tendency Ranges =Max(cell 1: cell N) =Min(cell 1: cell N) Percentages Frequency tables Conditions

21 Basic Calculations Totals Middle tendency Ranges Percentages =(Cell 1/Sum)*100 Frequency tables Conditions

22 Basic Calculations Totals Middle tendency Ranges Percentages Frequency tables Helps identifying how things are distribute within the data Conditions

23 Basic Calculations Totals Middle tendency Ranges Percentages Frequency tables =Countif(Cell 1: Cell N, value ) Conditions

24 Basic Calculations Totals Middle tendency Ranges Percentages Frequency tables Conditions Great to identify the any of the above by one or more populations

25 We Are Interested in Six Tracts

26 Let s Answer Our Questions How long have households been in their current house? Which neighborhoods are most transient? Table 1. Percent of Households by Move-In Period and Census Tract Move-In Period Tract or Earlier

27 Let s Answer Our Questions We are going to fill in this table with a VLOOKUP formula. This finds records in a table where the IDs match. So we can use it to find the value we need for each tract ID. =VLOOKUP($A4,'Move-In Period'!$A$1:$G$225,2) What ID do you want to look up? Where is the table? Which column has the value you want to show?

28 Let s Answer Our Questions How long have households been in their current house? Which neighborhoods are most transient? Table 1. Percent of Households by Move-In Period and Census Tract Move-In Period or Earlier Tract

29 The bulk of households moved in between 2000 and Tract 3562 is the most transient, in that it has the most new residents. Table 1. Percent of Households by Move-In Period and Census Tract Move-In Period or Earlier Tract

30 Let s Answer Our Questions Which neighborhood has the highest number of rental households? Percent rental households? Table 2. Owner and Renter Counts and Rates by Census Tract Tract Pct. Owner Pct. Renter # Owner # Renter

31 Let s Answer Our Questions We are going to fill in this table with a VLOOKUP formula. This finds records in a table where the IDs match. So we can use it to find the value we need for each tract ID. =VLOOKUP($A25,'Owner Renter'!$A$1:$F$225,4) What ID do you want to look up? Where is the table? Which column has the value you want to show?

32 Let s Answer Our Questions Which neighborhood has the highest number of rental households? Percent rental households? Table 2. Owner and Renter Counts and Rates by Census Tract Tract Pct. Owner Pct. Renter # Owner # Renter

33 Tract 3569 has the highest number of rental households. Tract 3562 has the highest percent rental households. Table 2. Owner and Renter Counts and Rates by Census Tract Tract Pct. Owner Pct. Renter # Owner # Renter

34 This brings up a few more Questions. Tract 3562 stands out as having a high rental rate compared to this area. What about compared to the whole city? Tract 3562 had a lot of new households move in Is this because it has such a high rate of rentals?

35 To answer these, start by imagining your table. Tract 3562 stands out as having a high rental rate compared to this area. What about compared to the whole city? Table 3. Renter Rates by Area Pct. Renter Pct. Owner Tract 3562 Near Southeast Side Median Indianapolis Neighborhood

36 Then fill the table in. We already know the values for Tract Table 3. Renter Rates by Area Tract 3562 Pct. Renter 63.8 Pct. Owner 36.2 Near Southeast Side Median Indianapolis Neighborhood

37 Then fill the table in. Now let s use the sum formula for the Near Southeast Side. Add up owners, then add up renters. Now take renter/(renters+owners) and owners/(renters+owners). Table 3. Renter Rates by Area Tract 3562 Pct. Renter 63.8 Pct. Owner 36.2 Near Southeast Side Median Indianapolis Neighborhood

38 Then fill the table in. Now let s use the sum formula for the Near Southeast Side. Add up owners, then add up renters. Now take renter/(renters+owners) and owners/(renters+owners). Table 3. Renter Rates by Area Near Southeast Tract 3562 Side Pct. Renter Pct. Owner Median Indianapolis Neighborhood

39 Then fill the table in. To compare to Indianapolis, let s use the median neighborhood in terms of percent renters. =MEDIAN('Owner Renter'!F:F) Table 3. Renter Rates by Area Near Southeast Tract 3562 Side Pct. Renter Pct. Owner Median Indianapolis Neighborhood

40 Then fill the table in. To compare to Indianapolis, let s use the median neighborhood in terms of percent renters. =MEDIAN('Owner Renter'!F:F) Table 3. Renter Rates by Area Near Southeast Side Median Indianapolis Neighborhood Tract 3562 Pct. Renter Pct. Owner

41 Let s imagine the next table. Tract 3562 had a lot of new households move in Is this because it has such a high rate of rentals? Table 4. Pct. Of Households that Moved In Since 2015, by Rental Rate Above Marion Co. Rental Rate At or Below Marion Co. Rental Rate HHs Moved In Since 2015 All Households Pct. HHs Moved In Since 2015

42 Let s imagine the next table. We want to add up all households moved in since 2015 and total households for tracts above Marion County s rental rate, which is 46.4%. Table 4. Pct. Of Households that Moved In Since 2015, by Rental Rate Above Marion Co. Rental Rate At or Below Marion Co. Rental Rate HHs Moved In Since 2015 All Households Pct. HHs Moved In Since 2015

43 Let s imagine the next table. Our formula should read like: Sum(Count of HHs Moved in 2015) if Percent Rental > 46.4 Table 4. Pct. Of Households that Moved In Since 2015, by Rental Rate Above Marion Co. Rental Rate At or Below Marion Co. Rental Rate HHs Moved In Since 2015 All Households Pct. HHs Moved In Since 2015

44 Let s imagine the next table. Our formula will look like this: =SUMIF(percent rental, above 46.4",HHs moved in 2015) Values you want to test Criteria for the test Values you want to add, if this row passes the test

45 Let s imagine the next table. In our Clean Data tab, percent rental is in column CQ and percent moved in 2015 is in column CX. =SUMIF(CleanData!CQ:CQ,">46.4",CleanData!CX:CX) Values you want to test Criteria for the test Values you want to add, if this row passes the test

46 Let s imagine the next table. Now repeat, but add up total households for tracts that pass the test. =SUMIF(CleanData!CQ:CQ,">46.4",CleanData!CV:CV) Values you want to test Criteria for the test Values you want to add, if this row passes the test

47 Let s imagine the next table. Repeat these formulas for the At or Below Marion Co. Rate Column by changing the criteria of the test to: <=46.4

48 Let s imagine the next table. So tracts that have higher than average rental rates have about twice the proportion of households moving in 2015, but Tract 3562 is still much higher. Table 4. Pct. Of Households that Moved In Since 2015, by Rental Rate Above Marion Co. At or Below Marion Rental Rate Co. Rental Rate HHs Moved In Since All Households Pct. HHs Moved In Since % 4.5%

49 Charts

50 Let s show what we found in a chart What would be a good chart type show the renter/owner mix in Fletcher Place? What would be a good chart to compare this mix to the Near Southeast Side and to Indianapolis? What would be a good chart to compare Fletcher Place to the data we found in the last table (Pct. Moved in 2015 by Rental Rate)?

51 Renter/Mix in Fletcher Place Most Fletcher Place Households Are Renters Pie charts are great for percent breakdowns, as long as they are discreet groups that don t overlap. Don t name your chart what it is, name it what it means. Hit your audience over the head with your conclusion. Pct. Renter Pct. Owner

52 Comparing Fletcher Place to Area and City Fletcher Place Near Southeast Side Median Indy N hood Pct. Renter Pct. Owner Pct. Renter Pct. Owner Pct. Renter Pct. Owner

53 Comparing Fletcher Place to Area and City Fletcher Place Near Southeast Side Meh this is Median Indy N hood just okay Pct. Renter Pct. Owner Pct. Renter Pct. Owner Pct. Renter Pct. Owner

54 Comparing Fletcher Place to Area and City Fletcher Place Near Southeast Side Meh this is Median Indy N hood just okay Pct. Renter Pct. Owner Pct. Renter Pct. Owner Pct. Renter Pct. Owner

55 Pie charts are not great for comparing across categories. Think about what you really want to say, and then ask How can I say it most directly? For example, we want to say that Fletcher Place has a higher rental rate than the Near Southeast Side, and much higher Indy overall. So let s just make a chart of rental rates.

56 Pie charts are not great for comparing across categories. Think about what you really want to say, and then ask How can I say it most directly? For example, we want to say that Fletcher Place has a higher rental rate than the Near Southeast Side, and much higher Indy overall.

57 So let s just make a chart of rental rates. 70 Pct. Renter Fletcher Place Near Southeast Side Median Indianapolis Neighborhood

58 Add data labels, and now the vertical axis is unnecessary Pct. Renter Fletcher Place Near Southeast Side Median Indianapolis Neighborhood

59 The grid lines add clutter, and don t contribute meaningful info Pct. Renter Fletcher Place Near Southeast Side Median Indianapolis Neighborhood

60 Bar labels are too small, not readable Pct. Renter Fletcher Place Near Southeast Side Median Indianapolis Neighborhood

61 Let s give it a title that means something. Fletcher Place has a higher rate of rental households than the city or the Near Southeast Side Fletcher Place Near Southeast Side Median Indianapolis Neighborhood

62 Maps!

63 We are going to look at a few kinds of maps. Choropleth: Polygons are colored by indicator value Dot density: Polygons are filled with randomly placed dots according to indicator value

64 There a couple ways to color a choropleth map. Sequential Scale Diverging Scale Qualitative Scale Let s look at to see these in action.

65 There a couple ways to color a choropleth map. Sequential Scale Diverging Scale Qualitative Scale Let s look at to see these in action.

66 Choose an indicator and a legend, and make your own map. Indicators are on handout. Legends could be: Diverging scale Sequential Scale Qualitative Scale Dot Density