We need better data–Stat!

So let’s get the problems out of the way first:

– Our counts are not evenly distributed across 1875 to 2014; we have one cluster of dates from the cemetery’s dedication in 1889 to 1920, and one cluster running roughly from 2001 to 2014. Between these clusters, it’s kind of a crapshoot.

– We should have planned ahead to randomize our sampling across the cemetery, but it was not to be (i.e. we didn’t think to).

– The sample size is very small.

– There are so many reburials in this cemetery, but they can be difficult to identify, and we do not have perfect knowledge of the site. Reinterment is likely the biggest external confounding factor for our seriation.

But still! We forge on, like true statistics-doers, eager to see what sort of vague, baseless pattern we can squeeze out from terrible data. To investigate the question of changes in title over time, I made a stacked bar chart as well as a stacked area chart—seeing both of them together makes them somewhat more useful for this rudimentary depiction of seriation (Figs. 1 and 2 – to be inserted later)

By far, most markers of men or women bear no reference to familial titles.  For traditionally men’s names, if there is any identifying information besides the name, extra-familial duty can be emphasized: reverends, cops, individuals who worked in the military can be marked in some way, either by simply listing their title, battalion, etc., or by a logo/insignia inscribed on the marker, but familial roles like “Father”, “Husband”, etc. are far less common and appear in far less variety than they do for traditional women’s names.

For these identified women, familial roles are far more emphasized. If there is any inscription besides the name and the dates of life, then the inscription usually includes “Wife”, “Mother”, or (less common) “Sister”. Other flights of fancy like “Mom” or “Nana” are also seen, especially more recently. We found two examples of markers that bore 4 or more familial titles for one person.

Perhaps this is indicative of a true pattern, but it is doubtful. Sure, the data is in line with our perceptions of how men and women were and are perceived, but the data are so poor it’s difficult to say much of anything.

 

4 thoughts on “We need better data–Stat!

  1. Interesting consideration regarding the Catholic cemetery and potential double-meanings of kinship terms, Lauryl. I’ve generally seen priests and nuns buried in distinct cemeteries associated with their orders (e.g., Franciscan, Carmeltie, etc.) or with the church they’ve served.

    Red0, how many more graves would you need to survey in order to begin to make generalizations about the burials? I’m curious here to see how long it took your team to record all the headstones and figure out if there’s a way we could a more systematic survey that compiles data from multiple teams.

    • Indeed, our team has the similar problem, we divided into three survey route and try to cover the cemetery holistically and efficiently at the same time. But according to our survey map (https://mapsengine.google.com/map/edit?mid=zxFpUVURqYaE.k_IY00Zf0kto), to conduct a survey without the original survey bias is hard. All three of us record too many gravestone in the beginning, and I did not go the the southeast end of the cemetery, so we basically have no data over there….

  2. We had very similar problems- mostly due to the large size of the cemetery we all sort of stuck together in the southwest portion of the property by chance. Interestingly, we found that husband/father were roughly distributed in the same proportion as wife/mother. In fact, there were some gravestones that I saw that that were written by the wife (e.g. “beloved husband of xyz”). But since this is a Catholic cemetery, it can be pretty confusing, because father and sister are both familial terms as well as religious positions, and it wasn’t always clear whether the epitaph was marking someone’s Catholic order, or actually denoting kinship.

Leave a Reply

Your email address will not be published. Required fields are marked *