Data dictionary and design

Broiler feed efficiency

Synthetic teaching data. All observations are invented, not the source paper’s raw data.

Researchers house 20 broilers in each of 32 pens. They group the pens into eight sets of four pens in similar locations; these sets are called blocks. Within each block, they use a random order to assign one pen to each diet: control, BSF, LMW or YMW. All birds in a pen receive the same diet for 35 days. Researchers measure total feed disappearance and total liveweight gain for each pen over that same period. The invented study includes no deaths or removals.

Researchers assign diet to each whole pen: eight pens per diet and 32 pens in total. Measuring 640 birds does not turn this into 640 separate diet assignments.

Column	Unit/type	Meaning
pen_id	ID	Bpen01–Bpen32; one row per pen
block	category	block01–block08; four pens/diets in each block
diet	category	control, BSF, LMW or YMW; assigned to whole pens
birds	birds	20 per pen throughout the invented trial
feed_kg	kg	total feed disappearance for this pen over 35 days
gain_kg	kg	total liveweight gain for this pen over the same 35 days
fcr	kg feed/kg gain	feed_kg/gain_kg; continuous pen outcome

No values are missing; NA is the missing-value convention if a group adds external data. IDs are invented and contain no personal information. Rows are complete by construction; this does not imply missingness is harmless in a real experiment. No weights, outcomes or sample sizes were calibrated to published results.