## 1690

 Pro-Parris Tax Anti-Parris Tax Mean 14.76 Mean 15.82 Median 14.00 Median 14.00 Mode 7.00 Mode 15.00 Range 37.00 Range 46.00 Minimum 3.00 Minimum 4.00 Maximum 40.00 Maximum 50.00 Sum 561.00 Sum 538.00 Count 38.00 Count 34.00

The 1690 Tax Data Set can be used to compare the relative wealth of those associated with the pro- and anti-Parris factions by using descriptive statistics, sometimes called summary statistics. Descriptive statistics simplify and summarize information. In particular, for our purposes, they describe the central tendency of data: the average (mean) tax; the median tax, which is the middle tax (half the numbers are equal to or above, and half are equal to or below this value); and the mode, which represents the tax occurring most frequently. Each measure shows a different aspect of a group's taxes. For example, the median tax compensates for a few cases of extremely high or extremely low taxes that may distort the picture of a group's condition.

To describe and compare the pro- and anti-Parris taxes, select the data set and, from the Data menu, apply the "AutoFilter" to all columns. Click the autofilter icon on the "Petition" column and select "Anti-P." From the Tools menu, select "Data Analysis" and "Descriptive Statistics." Click the "Input range" button and drag through the column (include label) of the Anti-P taxes. Select the "Columns" button and the "Labels in first row" and "Summary Statistics" boxes in the "Descriptive Statistics" window. Place the output in a new worksheet. Repeat the process for the "Pro-P" group. The two tables can then be edited (as above) and compared.

The data confirm the idea that Parris's enemies were wealthier than his supporters: the average tax of the anti-Parris group was 15.8 shillings compared to 14.8 for Parris's supporters. Moreover, the highest (maximum) tax paid by an anti-Parris villager was ten shillings higher than the highest tax of a pro-Parris resident.

Nevertheless, the difference between the two groups was not large. Only one shilling separated them, considerably less than the amount that Boyer and Nissenbaum found by using a different tax list, one taken five years later (1695). Even more significant, the median tax for the two groups was identical at fourteen shillings, further suggesting that at the time of the witchcraft outbreak, the village's two factions were not widely separated in wealth.

The chart's "Count" row reveals some limitations to this analysis. Only thirty-eight of the fifty-four male church members and householders who signed the pro-Parris petition in 1695 were assessed taxes in 1690. On the anti-Parris side, only thirty-four of the forty male church members, freeholders, and householders who signed the anti-Parris petition were assessed taxes in 1690. Nevertheless, the evidence from approximately 70% of the pro-Parris petitioners and 85% of the anti-Parris petitioners casts doubt on the idea that disparities of wealth determined Salem Village's factionalism.

Finally, users can see how both factions compared to the tax profile of Salem Village as a whole by determining the descriptive statistics for the entire 1690 tax list.

We can also refine our investigation by examining different categories of taxpayers. Click Next.