Dairy cow dry matter intake for multiverse analysis and Bradley-Terry modeling
Data files
Jan 16, 2024 version files 8.29 KB
-
dairy_cow_dry_matter_intake_for_multiverse_and_bt.csv
6.39 KB
-
README.md
1.90 KB
Abstract
The palatability of feed for dairy cows is an important consideration but is difficult to measure, particularly when considering more than two feeds. We outline how a combination of multiverse analysis and Bradley-Terry modeling, two methodological tools that have rarely been applied in dairy science, can be adapted to address this problem. Specifically, we propose to apply multiverse analysis as a way to consider a range of thresholds for how much of a mixed grass-legume (MGL) silage had to be consumed (as a percent of the total DMI) to be designated as preferred. Each threshold gives rise to a separate dataset and a corresponding fitted Bradley-Terry model. Bradley-Terry models attribute to each feed what is commonly referred to as an “ability” in the context of sports or other competitions but can be interpreted as palatability when applied to feeds. This combined approach is a way of estimating palatabilities that appropriately reflect the degree of preference cows express through their feeding behavior. It has the advantages of being transparent and relatively easy to implement. A possible disadvantage is that this method is limited to a paired comparison approach and has difficulties with main-effects statistical inference. We demonstrate the use of this methodology on an example dataset comparing MGL silages under different ensiling conditions and exposed to oxygen for different durations.
README: Dairy cow dry matter intake for multiverse analysis and Bradley-Terry modeling
https://doi.org/10.5061/dryad.sn02v6xbd
This dataset originates from a study conducted at the William H. Miner Agricultural Research Institute (Chazy, NY) during summer 2021. Feed from two bunks (that is, two ensiling conditions) and from two aerobic exposure times (0 and 48 hours) were considered, for a total of four feed types. The four feed types were offered to twelve heifers in a pairwise fashion (4 types = 6 pairwise comparisons). The observations consist of the dry matter intake of each heifer by feed type and day. This dataset is being employed as a testbed for an statistical analysis approach consisting of multiverse analysis and Bradley-Terry modeling.
Description of the data and file structure
The dataset consists of a single CSV file (dry_matter_intake_for_multiverse_and_bt.csv). The unit of observation is a feed offered on a particular day to a particular cow. That is, each row is "half" of one pairwise comparison. There are 144 observations (12 heifers x 2 feeds/day x 6 days). The file contains nine columns:
date - date in M/D/YYYY format
cow.id - cow ID
trmt - feed type (A0 = bunk A, 0 hours aerobic exposure, A48 = bunk A, 48 hours aerobic exposure, B0 = bunk B, 0 hours aerobic exposure, B48 = bunk B, 48 hours aerobic exposure)
bunk - bunk the feed was taken from (A or B)
time - aerobic exposure time, in hours (0 or 48 hours)
pos - position in which the feed was offered to the heifer (L=left, R=right)
feedout.t.C - temperature (in degrees Celsius) of the feed at time of feeding
dmi.kg.30mins - dry matter intake (in kg) after 30 minutes
dmi.kg.3h - dry matter intake (in kg) after 3 hours
No observations are missing.
Code/Software
Self-documented code for applying the proposed methodology to the dataset is in "multiverse_and_bt_for_palatability.R".
Methods
Both experimental setup and statistical methodology are described in the accompanying publication. The accompanying R script also provides details of the statistical methodology.