Last updated: 2026-03-17

Checks: 1 1

Knit directory: HonorsThesis/

This reproducible R Markdown analysis was created with workflowr (version 1.7.2). The Checks tab describes the reproducibility checks that were applied when the results were created. The Past versions tab lists the development history.

R Markdown file: uncommitted changes

The R Markdown file has unstaged changes. To know which version of the R Markdown file created these results, you’ll want to first commit it to the Git repo. If you’re still working on the analysis, you can ignore this warning. When you’re finished, you can run wflow_publish to commit the R Markdown file and build the HTML.

Repository version: 3216d85

Great! You are using Git for version control. Tracking code development and connecting the code version to the results is critical for reproducibility.

The results in this page were generated with repository version 3216d85. See the Past versions tab to see a history of the changes made to the R Markdown and HTML files.

Note that you need to be careful to ensure that all relevant files for the analysis have been committed to Git prior to generating the results (you can use wflow_publish or wflow_git_commit). workflowr only checks the R Markdown file, but you know if there are other scripts or data files that it depends on. Below is the status of the Git repository when the results were generated: 


Ignored files:
    Ignored:    .DS_Store
    Ignored:    .claude/
    Ignored:    renv/library/
    Ignored:    renv/staging/

Unstaged changes:
    Modified:   Data/period_dictionary.csv
    Modified:   analysis/cups_and_skulls_analysis.Rmd
    Modified:   analysis/index.Rmd

Note that any generated files, e.g. HTML, png, CSS, etc., are not included in this status report because it is ok for generated content to have uncommitted changes.

These are the previous versions of the repository in which changes were made to the R Markdown (analysis/index.Rmd) and HTML (docs/index.html) files. If you’ve configured a remote Git repository (see ?wflow_git_remote), click on the hyperlinks in the table below to view the files as they were in that past version.

File Version Author Date Message
html d3ca59f tinalasisi 2026-02-03 Add data exploration analysis with Laconia burial data
Rmd 16128ad teamess 2026-02-03 Add jQuery UI docs, data files, and Rmd tweaks
html da0ef5a teamess 2026-01-27 Build workflowr website
Rmd 2d45cf5 teamess 2026-01-27 Start workflowr project.

Drinking Vessels, Skeletal Remains, and Hair Pins in Mycenaean Laconia and Messenia

This website presents the quantitative analyses underlying an honors thesis examining funerary assemblages from Bronze Age southern Greece. The dataset comprises tomb-level records from Laconia and Messenia spanning the Middle Helladic through Late Helladic periods (approximately 2000–1070 BCE). The analyses investigate whether drinking vessels, skeletal remains, and hair pins co-occur in patterned ways within and across these two regions, and whether those patterns shift over the course of the Bronze Age.

Methods

Dataset Construction

The dataset was compiled from two published regional burial catalogues: Death in Mycenaean Laconia: A Silent Place (Chrysanthi Gallou 2020) and Middle Helladic and Early Mycenaean Mortuary Practices in the Southern and Western Peloponnese (Michael J. Boyd 2002). Each row in the master spreadsheet (laconia_messinia_cups_skulls_pins.csv) corresponds to a single tomb or grave feature and records the site name, regional attribution, tomb type as described by the excavator, chronological period as assigned in the publication, estimated chamber area (where reported), burial practice (primary and/or secondary), the presence or absence of drinking vessels and a count of vessels where available, the presence or absence of identifiable human remains and a count of skulls where reported, and the presence or absence of hair pins. All data entry was performed by the author directly from the published sources; entries for which the published record was ambiguous or silent were recorded as missing (NA) rather than inferred.

Variable Coding

Raw text fields were standardized into analysis-ready variables. Drinking vessel presence, skeletal remains, and hair pin presence were each recorded in the spreadsheet as “yes”/“no” fields and converted to binary logical values (TRUE/FALSE); ambiguous or unrecorded entries were retained as NA and excluded from any calculation where the relevant variable was required. Primary and secondary burial were similarly recoded from “present”/“absent” fields.

Burial contexts were then classified into four mutually exclusive categories based on the combination of primary and secondary evidence: Primary Only, Secondary Only, Both Primary and Secondary, and Unknown/Not Recorded. Tomb types were simplified from their original descriptive labels into five structural categories — Tholos, Chamber Tomb, Cist Grave, Pit Grave, and Dromos Only — to permit comparison across sites. Graves that did not fall into these architectural categories were excluded from tomb-type analyses.

Chronological Classification

Period attributions in the source publications range in precision from a single sub-phase (e.g., “LHIIIA2”) to spans covering several centuries (e.g., “LHIIA–LHIIIB”). All period codes in the dataset were matched against a hand-compiled period dictionary (period_dictionary.csv) that records, for each code, its earliest and latest component phases and a chronological ordering number. Calendar date ranges were assigned based on the periodization chart in Tartaron (2007); see the data dictionary for specific date assignments.

For broad temporal analyses, each grave was assigned to one of seven phases — MH, LH I, LH II, LH IIIA, LH IIIB, LH IIIC, and Submycenaean — based on the earliest component of its recorded period. This is a conservative choice: a grave dated “LHIIA–LHIIIB” is counted in the LH II phase. It avoids placing graves in later phases they may not represent, but it means that widely dated graves are anchored at the earliest possible position. Graves whose period code could not be matched to any known component, or for which no period was recorded in the source publication, were excluded from all temporal analyses.

Statistical Methods

All statistical tests were chosen to match the structure of the data and the size of the available samples.

Frequency distributions and cross-tabulations are used throughout to describe the prevalence of drinking vessels, hair pins, and skeletal remains by tomb type, burial category, and region. Counts and within-group percentages are reported for all frequency tables.

Fisher’s exact test is used to test associations between pairs of binary variables (e.g., hair pin presence and secondary burial presence). Chi-square tests were not used because the sample of pin-bearing graves is small (n = 16 total across both regions), and several expected cell counts fall below the conventional threshold of five observations required for chi-square to be valid. Fisher’s exact test makes no assumption about minimum cell size and is therefore appropriate here. Results report the p-value and odds ratio.

Wilcoxon rank-sum test (Mann–Whitney U) is used to compare chamber area distributions between graves with and without hair pins. Chamber area values are right-skewed — a small number of very large tholos tombs pull the distribution — so a non-parametric test is preferred over a t-test, which assumes approximate normality. Results report the test statistic and p-value separately for each region.

Pearson correlation and ordinary least squares regression are used to examine the relationship between the number of drinking vessels recorded and the number of skulls recorded per grave, as a test of whether vessel deposition scales with the number of individuals interred. Regression results report the coefficient estimate, standard error, t-statistic, and p-value.

Missing data are handled consistently throughout: any observation with a missing value in a variable required for a given analysis is excluded from that analysis only, without affecting other calculations. The total number of observations contributing to each test is reported alongside the result.

Visualization

All figures were produced with ggplot2. Each chart reports raw counts alongside within-group percentages to allow direct comparison across groups of different sizes. Temporal plots are produced in two versions: one with equal spacing between broad phase categories (Option A), and one with bar widths scaled proportionally to actual phase duration in calendar years (Option B), so that the unequal lengths of successive phases are visually apparent.

Reproducibility and Data Availability

All analyses were conducted in R (version 4.5) using the workflowr package (Blischak et al. 2019), which manages the project as a version-controlled website in which each analysis page records the Git commit hash and working-directory status at the time it was built, making it possible to reproduce any result exactly. The full project — including raw data, analysis code, and rendered outputs — is publicly archived on GitHub. Package dependencies are managed with renv to ensure that the precise library versions used here can be restored. Key packages include dplyr and tidyr (data manipulation), ggplot2 (visualization), knitr and kableExtra (table formatting), and broom (tidy model output).

Site Contents

Reference

Blischak JD, Carbonetto P, Stephens M (2019). Creating and sharing reproducible research code the workflowr way. F1000Research 8:1749. doi:10.12688/f1000research.20843.1