2.4. Sample of animals and statistical analysis
Lehner (1992) proposed ‘sampling of all occurrences’ as the method of choice to generate accurate frequency and duration data. For our study, videotapes, each of 60 h length, from 438 litters were available. To realise this examination, a block data design of 40 animals was chosen. Twenty sows without any crushed piglets (NC-sows) were randomly sampled by using uniform distribution procedure (SAS, 2004). Sows with one or more crushed piglets within the first 48 h after parturition (C-sows) were selected conforming the NC-sow group (Table 3), considering the matching criteria number of piglets born alive, parity and farrowing date (season). Significance of differences between the means of NC- and C-sows was tested by using the TTEST and NPAR1WAY procedure of SAS (2004).