Statistical Analysis of Long and Short-run Equilibrium of Egg Hatchability in Two commercial hatcheries in Ibadan Metropolis

Abstract

The challenge of dearth of information on egg hatchability variables (EHV) for strategic planning can sufficiently be met using a statistical tool like cointegration which establish relationships between such egg hatchability variables (EHV) and considers the time series nature of the data. We conducted this study to evaluate the cointegration of EHV (egg set, fertile egg and total chicks hatched) in 2 selected commercial hatcheries in Ibadan Metropolis and to compare the results of the co- analysis from the 2 hatcheries. We adopted Engle and Granger co-integration tests method to assess long and short run equilibrium using data from two commercial hatcheries - Bronco and Foresight hatcheries at Oluyole Estate (Latitude 7º 23´ N and Longitude 3º 82´ E) in Ibadan. The mean egg hatchability variables were 5118.421, 4396.653, and 3796.335 for egg sets, fertile eggs and total chicks for farm 1. These were relatively lesser than 13076.160, 11717.280 and 10462.050 returned for egg set, fertile egg and total chicks for farm II. It was obtained that for each of the variable at both farm, the ADF is significant at 0.01, 0.05 and 0.1 level of significant. This implied that the egg hatchability variables (EHV) data and for both farms were non stationary. Similarly, they were non-random walk process because the data were still non-stationary after the first and second differencing. It was also observed that the unit root for the EHV, since the absolute values of EHV is greater than unity (- /B/>1), hence the non- stationarity of the EHV could be described as explosive. Three types of causality direction can be identified in this study – Unidirectional causality running from total chicks to egg set (at Farm II), bidirectional causality running from egg set to fertile egg (at both farm) and no causality between the 2 EHV as egg fertile and total chicks (at Farm II). Sum of square of residuals of VECM for total chick was the least while the adjusted R2 for the VECM ranged from 0.362 for fertile egg to 0.456 for egg set for farm I and for farm II, it ranged between 0.311 for total chicks and 0.449 for fertile egg. The implications of these results were that long run equilibrium could be established between all EHV. We conclude from our study that relationship between EHV in this study is a genuine relationship arising from the cointegration of the EHV and not a spurious regression.