Social Explorer paper draft instructions:

(draft instructions updated May 9, 2022)

* papers should be short: 2-4 pages of text plus 1 or 2 figures generated by Social Explorer. In our comments on your drafts, we may ask you to add more figures, so you may end up with 3 or 4 pairs of figures.

Social Explorer:

http://www.socialexplorer.com/pub/maps/home.aspx

Follow the link above, select Product> Maps and hit the “Dashboard” button.

Any modern browser should work. [Old note: Social Explorer offers a flash and a non-flash option. If you use the flash option (which I do *not* recommend) you *must* have Adobe Flash enabled on the browser (try Internet Explorer)- I am not sure if this is still true]

The little dot pattern in the upper left will take you back to your dashboard home page.

I suggest that you follow the "US Demography" link, as the US Demography maps are most relevant to our class, but you may want to check out the US election data link. The Ireland and UK map links are out of bounds for this class. Use the Decennial Census data sources.

Set up a free account on Social Explorer and save your work as you go. When you go to Dashboard, you should see your saved projects there. Remember to save frequently (or save under a new name) because it is not so easy to “undo” actions in Social Explorer.

Two Figures that show the same phenomenon at different at different census years is good strategy (but not the only strategy).

If you want to show change over time, it is usually helpful to pick two census years that are far apart, like 1960 and 2010. On the other hand, if all the change happened between 1940 and 1950, the close years’ comparison would be more persuasive as to what the cause of the change was.

One new feature of recent versions of Social Explorer is that you can generate tables in addition to maps. If you want to try to generate an index of dissimilarity or an index of isolation, you could use the Table function to generate an Excel file, and then use the formulas from my web-posted Excel file to generate the indices, but note: this is not required and if you find yourself getting stuck doing this, go back to the maps. The maps are what this project is about!

* Note that census income is in nominal dollars (not inflation corrected), so $20,000 household income was rich in 1940, but working poor in 2010. Percent of households in poverty is reasonably consistent over time (because the income that is used in the poverty measure is indexed to inflation). Newer note: Social Explorer can correct for inflation for you! Any year’s data, including maps of income, rent, or home values can be adjusted for inflation to any other year, so you don’t have to do any clunky hand adjustments. See their instructions here:

https://www.socialexplorer.com/blog/post/social-explorer-debuts-new-features-9591#feature6

https://www.socialexplorer.com/help/using-maps/visualizing-data/adjust-values-for-inflation-in-maps

* Note the Social Explorer facility for linking the geography of two maps side by side, which will also give you the readings for both values for the same census tract, which may be interesting.

* Cutpoint comparability is important: if you are looking at two census years side-by-side, and if you are looking at numbers or percentages that are comparable over time (such as percentage black, or percent in poverty, i.e. not dollar values that would have to be inflation-adjusted to be comparable), then you should adjust the cutpoints of the two maps to be the same, to make the two maps as comparable as possible.

* The cutpoints also should be set to values that are reasonable for the distribution of values in your map. For instance, if you are mapping population density, the cutpoints for most of the country (where the default highest category might be 15,000 persons per square mile) is no use in New York, where most census tracts have more than 100,000 persons per square mile. The “cutpoints” tool below the arrow next to the legend will show you a histogram of values for your particular map, which should let you set the cutpoints to reasonable values. If most of the map is the same two colors, your cutpoints are set incorrectly.

* And speaking of colors, the Social Explorer default color palette for "shaded area" maps is the palette with shades of brown, but those shades of brown are nearly impossible to distinguish, so if it is important for the reader to distinguish the values, think of using a palette with a more distinguishable set of colors (i.e., two colors or more rather than a monochrome map with shades of one color).

* Social Explorer allows 3 visualization types: “shaded area,” “bubbles,” and “dot density.” I have not myself found great uses for the “bubble” maps, but if you can find a good way to use them, that is great. The dot density maps are terrific for showing the geographical distribution and density of counts of things, for instance black population or white population or immigrant population or houses worth a certain dollar value (but see the above notes on nominal values). You could show where black people live with the dot density map, or percentage black by census tract with the shaded area map. Be as clear as you can be in your explanation of the maps, as to what is actually being shown.

* Dot maps can plot several populations on the same map, which is cool, but legibility requires that you make the dots substantially different colors so the different populations can be easily distinguished.

* When you put your mouse over each area in the Social Explorer map, a box with the actual data will pop up. Use the actual data, the counts, the dollar values, the numbers, in your paper as much as you can.

* Be cautious of the ecological fallacy. The ecological fallacy occurs when you infer things about individuals based on area means or medians. The Social Explorer maps give you area (for local maps, this is census tract) means or medians. Not everyone in the area has values equal to the area mean or median. You may be able to show that, for instance, Hispanics are over-represented in the neighborhoods with low household income. This suggests that Hispanics themselves have low income, but there are other possibilities, such as well-to-do Hispanic families living nearby to poor whites, because the well-to-do Hispanics are unable to buy in the exclusive or more well-to-do neighborhoods. You can certainly make inferences by comparing maps of race and income, but you should do so cautiously.

* National level maps are very difficult to judge because most of the US population lives in metropolitan areas that are tiny in size, whereas the big rural counties that are most visible in the national level maps have few people in them. I suggest you use maps at a more local level. If you use national maps you need to be cautious as to interpretation.

* In some situations it can be useful for the student to make their own annotations on the Social Explorer map, for instance labeling a key street, or where the high school and its district boundaries are, or circling the spot where a big housing project was built, etc. Social Explorer has a geographic Search function, but I have found Social Explorer’s geographic search function to be inadequate for identifying neighborhoods or schools or buildings, so if there is a particular place that is crucial to your story, you will need to identify your place of interest on the map within Powerpoint (or some other drawing program) after you produce the map in Social Explorer.

* Papers are electronic only, in Word or PDF format. I recommend that you submit your paper in Word format, because it is easier for us to put comments in the margins in your Word documents. Macs sometimes have trouble reading marginal comments in PDF files.

* Before you export your maps, try to enable the legends, or else you will have to add the legend yourself later (which dot is which population, for instance).

* I recommend that you export your Social Explorer maps image (.png) files using the highest resolution you can (not 96 dpi, but 192 dpi; 288 dpi is probably overkill), save the image to your local system, and then embed the files into your paper, preferably each on a separate page at the end, labeled Figure 1, Figure 2, etc. It is your responsibility to make the maps legible in the final paper. One reason to export your maps at high resolution is so that we can zoom in on your maps, and see details. If the exported image lacks a title or a legend, it will be important for you to add the title and legend by hand.

See this link on how to save your maps as images:

https://www.socialexplorer.com/help/using-maps/managing-maps/export-a-map-as-an-image

but note: if you have the cool side-by-side maps, the export to image button will ask you to pick one of the two maps to export. That is OK: you can present the two maps one above the other in your write up. Exporting as an image is your best option because it allows you to control the output resolution.

The other option is to save your map or series of maps as a “story” and then export it as a Powerpoint file. The problem with exporting to PowerPoint is that it is difficult to get the map at high resolution from PowerPoint to Word.

* You can also hit the little triangle on the lower right, for “tell a story” which will allow you to build a file with several maps, and export the file into Powerpoint. Powerpoint is a good program for overlaying arrows and text onto your maps, before copying the maps back to Word (Word also has drawing tools for overlaying arrows and text onto graphics). Also note: The PowerPoint file you export from your Social Explorer Story may have all the resolution you need, but when you copy the slide from PowerPoint to Word or your text editor of choice, you may find that the map has lost resolution!

You can try to change the output resolution of PowerPoint (but I don’t recommend this for Windows and I don’t know how to do this for Macs). So: exporting maps one by one each as an image is the better choice.

* Note: if Social Explorer truncates the titles or legend of figures, it is your responsibility to add titles above the figures that very clearly explain what the figure is. Also note: you may want to set the MS Word pages with the maps to be landscape rather than portrait orientation. You can accomplish this by inserting a section break between the text and the maps, and then changing the page orientation of the map area.

* Final papers (Still 2-4 pages plus several figures) are due to be uploaded to Coursework. Don't over-write.

* The paper should probably reference something in the literature from the class, but the paper should be mainly about the observed data, and only secondarily about the literature in the class.

* Your figures need to have notes below them indicating which data source, which year, and which variables were used.

* Your final paper should take the feedback you received on the draft version into account.

* Take a close look at the papers from past students posted to Canvas, for examples of how to use the maps and how to write about the maps.