The objective of this study was to validate scan sample accuracy at various predetermined intervals relative to continuous observation of individual nursery pigs interacting with a feeder and nutritional enrichment. A total of 280 mixed-sex pigs, 19 to 24 days of age were randomly allocated to 28 pens. Pen was the experimental unit, and 7 pens were assigned to each of 4 treatments: (1) biscuit with maternal pheromone attractant (MP), (2) biscuit with sugary attractant (strawberry jam) (JAM),(3) biscuit with no attractant (positive control; POS), and (4) no biscuit (negative control; NEG). Each enrichment pen received 4 biscuits suspended from 2ropes at the feeder for the first 7 days after weaning. The biscuit enrichments were added to the nursery pen at ~ the 15:00 hour on Day 0 and ~ the 08:30 hour on Days 1-6. Behavior observations were performed using continuous sampling of all nursery pigs in each pen during the first 30 minutes after biscuit placement on Days 0 to 6 (98 total hours, 3.5 hours/pen). The number of time each pig interacted with the feeder and biscuit was recorded as well as the duration of each interaction. These continuous observations were then used to create different data sets at varying sample times. The scan sample and continuous data were analyzed using the GLM procedure of SAS and behavioral frequency data were expressed as percentage of time (s) pigs were performing an identified behavior during the 30-minute observation periods. For enrichment use, there was no observed statistical difference between data collected using continuous observation, 10 second scans and 15 second scans when comparing continuous to 5 second scans, 10 second scans, 15 second scans and 30 second scans. For feeder use, there was no observed statistical difference between data collected using continuous observation and the 30 second scan sample when comparing continuous to 15 second scans, 30 second scans, 45 second scans and60 second scans. In conclusion, 30 second scan sampling for feeder use and 10second and 15 second scan sampling for enrichment use can be used, without being different to the continuous or “gold  standard” methodology. These scan point validations are critical because they allow the observer to collect more total data in less time without giving up accuracy.
Read full abstract