Joy through Statistics History Foundation Joy through Statistics 22/04/2019 History Foundation
Historians and Counting Economic historians do it Demographic historians have always done it Social historians began to do it in the 1960s Political historians have to do it Even educated students do it in dissertations It really isn’t difficult You don’t even have to count – Excel will do it for you 22/04/2019 History Foundation
What has pushed quantification? ‘Democratisation’ of history - history of the mass rather than of GREAT MEN - note the impact of the Annales school The ‘explosion’ of the social sciences in UK universities in the 1960s. The advent of cheap computing and cheap analytical packages Professionalisation of society 22/04/2019 History Foundation
Advantages of quantification Only with statistics can you study the masses Even with very limited processing, descriptive statistics can give greater accuracy than words Only with statistics is is possible to infer with any real confidence Using time series, it is possible to analyse with some confidence Elaborate analytical models can be tested 22/04/2019 History Foundation
Disadvantages of Quantification Data my not be robust or precisely what the historian needs Analytical techniques can be abused (Sir Cyril Burt) or be very sensitive to minor changes in the way data are ordered (UK Industrial Revolution) Quantification can only suggest relationships, the historian must interpret the results Historical models are very controversial (US RRs) 22/04/2019 History Foundation
An Example: Marriage in early 15th Century Tuscany Maristella Bottincini, ‘A loveless economy’, Journal of Economic History, 59 (1999) She asks: Did couples marry for love or was marriage a politico-economic union of families? 22/04/2019 History Foundation
What sources do Quantitative historians use? She worked in the State archives in Florence She used dowries registered between 1415 and 1436 Dowries were given by the bride’s family in marriage and were legal contracts She also used the Tuscan property census of 1427 22/04/2019 History Foundation
What questions do historians ask, Part 1 (characteristics of sample)? Males Females 18 54 14 31 22 30 15 23 27 24 26 16 17 25 What age were males and females at marriage? Use mean? (average - 27.8) How to cope with outliers? Use mode (commonest - 25) Use median (midpoint - 25) 22/04/2019 History Foundation
Step 1: enter the data into a spreadsheet 22/04/2019 History Foundation
Step 2: Decide on formulae 22/04/2019 History Foundation
Step 3: Let Excel calculate the results 22/04/2019 History Foundation
Marriage in 15th Century Tuscany, I Women tended to marry very young Female celibacy was virtually unknown Mean age of marriage for women was just under 19 Modal age was 16 Mean age for men was just under 28 Modal age was 25 22/04/2019 History Foundation
More refined descriptive statistics: marriage and social status Did daughters marry ‘up’ or ‘down’? She measures social status from the size of family wealth in the 1427 census and by occupation of father or groom Occupation is either in peasant agriculture or higher status non-agricultural work In her terminology agr and nonagr 22/04/2019 History Foundation
Marriage in 15th Century Tuscany, II 54% of all marriages were between peasant brides and peasant grooms; 14% between non-peasant brides and non-peasant grooms; the remaining 32% were ‘mixed’ A significant proportion (57%) of women ‘married down’ to grooms with lower family wealth Slightly fewer (43%) ‘married up’ to husbands of higher status or higher family wealth 22/04/2019 History Foundation
What questions do historians ask, Part 2 (relationships between variables)? Did parents use the dowry (given by the father of the bride, but controlled by the groom or his father) to ‘make matches’? Do parents act altruistically and ‘support’ daughters who marry ‘social inferiors’ by giving bigger dowries? Do they act opportunistically and seek ‘socially superior’ mates for their daughters by offering big dowries for good catches? 22/04/2019 History Foundation
Measures of association between variables Answers can be found to these questions by demonstrating whether there are statistically significant relationships between dowry size and social status through regression or correlation Data on dowry size is from the legal records & on the wealth of bride and groom families from the census of wealth Is there a statistical association between size of dowries and patterns of marriage? 22/04/2019 History Foundation
Step 1a: enter data into spreadsheet 22/04/2019 History Foundation
Step 1b: Complete data entry 22/04/2019 History Foundation
Step 1b (again): Complete data entry 22/04/2019 History Foundation
Step 2: Decide on formulae 22/04/2019 History Foundation
Step 3a: Let Excel calculate results 22/04/2019 History Foundation
Step 3b: Let Excel calculate results 22/04/2019 History Foundation
What about love and marriage? Dowries as compensation A daughter provides household service and work for her parents but consumes goods & services A wife provides similar household service and work and also consumes, but she also produces and raises children and cherishes her husband The daughter is thus a net cost to her parents & that cost mounts the longer she stays at home Do parents offer big dowries for older daughters? 22/04/2019 History Foundation
Love and Marriage: Correlation Coefficients Were positive between size of dowry and age of bride - it cost more to get rid of older daughters Were very positive between size of dowry and second marriage - widowed daughters were a big cost Were more weakly positive between dowry size and daughter who married down Were not positive when daughters married up 22/04/2019 History Foundation
Conclusions Ability to handle quantitative data is an essential tool for the historian Computerisation has democratised computation Botticini’s work shows that culture, attitudes and even family dynamics can be investigated using these methods Her analysis has richness and depth that can come only from a wide range of information 22/04/2019 History Foundation
Further Materials Mark Overton has prepared a number of basic statistical exercises available from the History Foundation web-site and will cover them in his lecture next semester This will act as a bridge to the best book on the introduction to quantitative techniques for historians, Pat Hudson, History by Numbers (2000). Review this lecture on http://www.ex.ac.uk/~aebooth/foundation 22/04/2019 History Foundation