Using factor scores in multiple linear regression model for predicting the carcass weight of

broiler chickens using body measurements

Uso de las puntuaciones del factor del eje principal en el modelo de regresión lineal múltiple para predecir el
peso de la canal de pollos de engorde usando medidas corporales
1
Abdulmojeed YAKUBU , Kingsley Omogiade IDAHOR 1 and Ya'u Isopa AGADE 2
1
Animal Science Department, Faculty of Agriculture, Nasarawa State University, Keffi, Shabu-Lafia Campus,
P.M.B. 135, Lafia, Nigeria and 2Department of Animal Science, College of Agriculture, Lafia, Nasarawa State,
Nigeria. E-mail: [email protected] Corresponding author

Received: 05/09/2009 First reviewing ending: 13/11/2009

First review received: 30/11/2009 Accepted: 28/12/2009

ABSTRACT

Information on the relationships among pre- and post-slaughter traits of broiler chickens is valuable to poultry farmers and
researchers as it allows early selection, as well as giving a chance to make an early evaluation of the breeding programme.
In this study, factor and multiple regression analysis were combined to examine the relationship between carcass weight and
six body measurements (live weight, body length, breast circumference, thigh circumference, shank length and wing length)
of 8-week old Arbor Acre broiler chickens in a sub-humid environment of Nigeria. In the varimax rotated factor analysis,
two factors were extracted which explained 87.53% of the total variability in the body measurements of chickens. The first
factor, termed Form Factor, had its loadings for body weight, body length, breast circumference and thigh circumference
respectively. The second factor was characterised by shank length and wing length, and was thus referred to as Length
Factor. These two factors together accounted for 84.50% of the variation in carcass weight when their scores were
considered as independent variables in the fitted multiple linear regression equation.

Keywords: Broilers, body measurements, factor analysis, rotation, regression, multicollinearity.

RESUMEN

La información sobre las relaciones entre los caracteres pre-y post- sacrificio de pollos de engorde es muy valiosa para los
avicultores e investigadores, debido a que permite la selección temprana, así como da la oportunidad de hacer una
evaluación precoz del programa de mejoramiento. En este estudio, la metodología del factor del eje principal y el análisis de
regresión múltiple se utilizaron para examinar la relación entre el peso de la canal y seis medidas corporales (peso vivo,
longitud corporal, circunferencia del pecho, circunferencia del muslo, longitud de tarso y longitud del ala) de pollos de
engorde Arbor Acre de 8 semanas de edad en un ambiente subhúmedo de Nigeria. En el análisis de factores rotados por
varimax, se extrajeron dos factores que explicaron 87,53% de la variabilidad total de las medidas corporales de los pollos.
El primer factor, denominado factor de forma, tuvo sus cargas para peso corporal, longitud corporal, circunferencia del
pecho y circunferencia del muslo, respectivamente. El segundo factor se caracterizó por la longitud del tarso y la longitud
del ala y fue denominado factor de longitud. Estos dos factores representaron 84,50% de la variación en el peso en canal
cuando sus valores se consideraron como variables independientes en la ecuación ajustada de regresión lineal múltiple.

Palabras clave: Pollos de engorde, medidas corporales, análisis factorial, rotación, regression, multicolinearidad.

INTRODUCTION determinants of the overall size and shape of a live

bird or carcass. According to Pinto et al. (2006), body
As with all animal species, information on weight at various ages and carcass characteristics are
correlations among the pre- and post- slaughter traits examples of variables that can indicate the usefulness
is quite important in poultry breeding. This is because of the chicken for commercial purposes.
knowing which of the pre-slaughter trait(s) affect
which of the post-slaughter trait(s) enables breeders to Performance testing, which forms the basis
predict what kind of product(s) will be obtained for breeding work is difficult to conduct in the case of
(Mendeş et al., 2005; Mendeş, 2009). Size and shape slaughter value parameters. Selection towards
(conformation) of various body parts are the major meatiness improvement requires reliable and easy-to-

Revista UDO Agrícola 9 (4): 963-967. 2009 963

 Yakubu et al. Using factor scores for predicting the carcass weight of broiler chickens using body measurements

apply methods for estimating the performance and taken: live weight (LW), body length (BL), breast
breeding value of poultry species (Kleczek et al., circumference (BC), thigh circumference (TC), shank
2007). Body measurements and meatiness traits are length (SL) and wing length (WL). The anatomical
inter-correlated (Shahin, 1999; Isiguzar, 2003). reference points were as described by Gueye et al.
However, the analysis of these traits should address (1998) and Teguia et al. (2008). For carcass
interdependence among the predictors evaluation, the birds were slaughtered by severing the
(multicollinearity). The problem in the analysis of carotid arteries and jugular veins and blood drained
body measurements and carcass weight data is the under gravity; scalded to facilitate plucking and
difficulty in interpreting the influence of body eviscerated. The carcasses were then divided into
measurements and determining the measurements parts as described by (Kleczek et al., 2006). The
which are most useful for predicting carcass weight weights of the thigh, breast and back were taken as
(Keskin et al., 2007). Hence, the use of a multivariate the carcass weight.
technique called factor analysis. This helps to uncover
the latent structure of a set of variables. Factor Means, standard deviations and coefficients
analysis reduces a large number of variables to a of variation of carcass weight and body measurements
smaller number of factors for modelling purposes of birds were calculated. Pearson’s correlation
(Tabachnick and Fidell, 2001; Manly, 2005). This coefficients (r) among carcass weight and the various
involves the use of factor scores for orthogonalization body measurements of birds were calculated. From
of predictors, thereby handling multicollinearity in the correlation matrix, data were generated for factor
such procedures as multiple regressions (Grice, analysis using principal axis factoring. The
2001). determinant of the correlation matrix was used to test
for multicollinearity and singularity. Anti-image
In Nigeria, the estimation of carcass weight correlations, Kaiser-Meyer- Olkin measures of
from body measurements of chickens using factor sampling adequacy and Bartlett’s Test of Sphericity
analysis has not been exploited. The present (tests the null hypothesis that the original correlation
investigation therefore, aimed at describing matrix is an identity matrix) were computed to test the
objectively the interrelationships existing between validity of the factor analysis of the data set. The
carcass weight and body measurements of broiler appropriateness of the factor analysis was further
chickens and to predict carcass weight from the tested using communalities and ratio of cases to
orthogonal traits derived from principal axis variables. Cumulative proportion of variance criterion
factoring. This will further aid in the selection and was employed in determining the number of factors to
breeding programmes of broilers in a sub-humid extract. Reproduced and residual correlations were
tropical environment. used to test the validity of the number of factors
extracted. The orthogonal varimax rotation was
MATERIALS AND METHODS employed to enhance the interpretability of the factor
loadings. Principal axis factoring is a method which
Data were obtained from one hundred and tries to find the lowest number of factors which can
twenty randomly selected unsexed 8-week old Arbor account for the variability in the original variables
Acre broiler chickens intensively reared at the Poultry that is associated with these factors (Wood et al.,
Unit of the College of Agriculture, Lafia, Nasarawa 1996; Rencher, 2002; Manly, 2005).
State, North Central Nigeria. It is located between
latitude 070 52’N and 08056’N and longitude 070 25’E If the observed variables are X1, X2,…, Xn,
and 09031’N respectively. The birds which were wing the common factors are F1, F2,…,Fm and the unique
banded for identification purpose were reared on deep factors are U1, U2,…, Un, the variables may be
litter from day-old. A finisher diet containing expressed as linear functions of the factors:
approximately 20% crude protein and a metabolizable
energy of 2900 kcal/kg was given from 5-8 weeks of X1= a11F1 + a12F2 + a13F3 + ... + a1mFm + a1U1
age ad libitum. Fresh clean water was also freely X2= a21F1 + a22F2 + a23F3 + ... + a2mFm + a2U2
available. Routine vaccination and other management …
practices were strictly adhered to. Xn= an1F1 + an2F2 + an3F3 + ... + anmFm + anUn

Prior to slaughtering the birds at weight 8 Each of these equations is a regression

weeks of age, the following measurements were equation; factor analysis seeks to find the coefficients

964 Revista UDO Agrícola 9 (4): 963-967. 2009

 Yakubu et al. Using factor scores for predicting the carcass weight of broiler chickens using body measurements

a11, a12,…,anm which best reproduce the observed Pairwise correlations among carcass weight
variables from the factors. The coefficients a11, and body measurements of birds are presented in
a12,…,anm are weights in the same way as regression Table 2. Carcass weight was positively and highly
coefficients (because the variables are standardized, correlated with the pre-slaughter traits investigated
the constant is zero). (r= 0.70-0.91; p<0.01). The correlation coefficients
between the body measurements ranged from
Factor scores were considered as moderate (r = 0.48) to high (r = 0.86) values
independent variables for predicting the carcass (p<0.01). Similar findings in other breeds/strains of
weight of birds using the following multiple chickens have been reported (Musa et al., 2006; Yang
regression model: et al., 2006; Ojedapo et al., 2008). In a related study,
Wolanski et al. (2006) reported that the relationship
CW= a + b1FS1 + b2FS2 +... + bkFSk + e existing between hatch weight, hatch body length,
hatch shank length and carcass weight was high and
Where, significant. The high association observed indicates
CW = carcass weight that carcass weight can be predicted from body
A = regression constant. measurements.
b’s = regression coefficients
FS’s = factor scores The determinant of the correlation matrix
e = random error term (0.001) was greater than 0.00001 while anti – image
correlations computed showed that partial correlations
The significance of the regression were low, indicating that true factors existed in the
coefficients was tested with a t-statistic while the data. This was buttressed by Kaiser-Meyer-Olkin
goodness-of-fit of the regression was assessed using measure of sampling adequacy studied from the
the coefficient of determination (R2) and adjusted R2. diagonal of partial correlation, revealing the
SPSS statistical package program was used to analyze proportion of the variance in the body measurements
the data (Anonymous 2001). caused by the underlying factor. The present value of
0.80 was found to be sufficiently high for all the body
RESULTS AND DISCUSSION traits. The overall significance of the correlation
matrix tested with Bartlett’s Test of Sphericity for the
The means, standard deviations and body measurements of birds (Chi-square = 776.90; p<
coefficients of variation of body measurements (live 0.01). The value of Bartlett test implied that the factor
weight, body length, breast circumference, thigh analysis is applicable to data sets. The communalities,
circumference, shank length and wing length) and which represent the proportion of the variance in the
carcass weight are presented in Table 1. Variability original variables that is accounted for by the factor
was higher in thigh circumference. This might not be solution ranged from 0.753-0.987. This further lent
unconnected with the high influence of environment
on these traits. Table 2. Phenotypic correlations (Pearson’s correlations)
between body measurements and carcass weight
of Arbor Acre broiler chickens intensively reared
Table 1. Descriptive statistics of body measurements and at the Poultry Unit of the College of Agriculture,
carcass weight of Arbor Acre broiler chickens Lafia, Nasarawa State, North Central Nigeria **.
intensively reared at the Poultry Unit of the
College of Agriculture, Lafia, Nasarawa State, Traits BL BC TC SL WL CW
North Central Nigeria. LW 0.85 0.80 0.71 0.79 0.70 0.91
BL - 0.70 0.65 0.62 0.48 0.73
Traits Mean SD CV BC - - 0.86 0.72 0.55 0.80
Live weight (kg) 1.76 0.26 14.77 TC - - - 0.70 0.69 0.71
Body length (cm) 36.74 3.89 10.59 SL - - - - 0.78 0.79
Breast circumference (cm) 32.15 4.74 14.74 WL - - - - - 0.70
Thigh circumference (cm) 11.36 2.60 22.89
Shank length (cm) 8.62 1.17 13.57 LW: Live weight (kg), BL: Body length (cm), BC: Breast
Wing length (cm) 18.58 2.25 12.11 circumference (cm), TC: Thigh circumference (cm),
Carcass weight (kg) 1.27 0.19 14.96 SL: Shank length (cm), WL: Wing length (cm) and
CW: Carcass weight (kg)
SD: Standard deviation; CV: Coefficient of variation (%) ** Significant at p<0.01 for all correlation coefficients.
Revista UDO Agrícola 9 (4): 963-967. 2009 965
 Yakubu et al. Using factor scores for predicting the carcass weight of broiler chickens using body measurements

credence to the appropriateness of the factor analysis. increase as the values of factor 1 and 2 scores
The ratio of cases to variables (17.3 to 1 far exceeded increase. A combination of these two independent
the minimum of 5 to 1 standard) was met as sample factors explained 84.50% of the total variability in
size requirement while estimates of the residual carcass weight. Similarly, Shahin (2000) and Shahin
correlation matrix were low enough. and Hassan (2000) derived regression equations for
estimating live weight of ducklings and rabbits
After varimax rotation of the factor axes, two respectively using independent factor scores. Keskin
factors were extracted which accounted for 87.53% of et al. (2007) also used factor scores derived from ten
the total variance of the original six variables (Table body measurements to predict the carcass weight of
3). Factor pattern coefficients of the rotated factors sheep.
show the relative contribution of each trait to a
particular factor. The first factor, which explained CONCLUSION
77.19% of the generalized variance was characterized
by high positive loadings (factor-variate correlations) The factor analysis aided in summarizing and
on live weight, body length, breast circumference and explaining the correlations and covariances among
thigh circumference. This factor was thus termed the original six body measurements of chickens: live
Form Factor. The variables that were more associated weight, body length, breast circumference, thigh
with the second factor were shank length and wing circumference, shank length and wing length. These
length, contributing to 10.34% of the variation. Thus, interdependent body traits were reduced to two
this factor was referred to as Length Factor. factors which were mutually orthogonal, thereby
eliminating multicollinearity problems among the
Factor score coefficients derived from the variables. Their factor scores fitted in a linear
body measurements of chickens are presented in multiple regression model revealed that these two
Table 4. The use of interdependent explanatory factors accounted for 84.50% of the variation in
variables should be treated with caution, since carcass weight of chickens. This is an indication that
multicollinearity has been shown to be associated factor scores of pre-slaughter traits could be
with unstable estimates of regression coefficients successfully used to predict post-slaughter trait such
(Ibe, 1989; Yakubu, 2009) rendering the estimation of as carcass weight, which could aid in selection and
unique effects of these predictors impossible. This breeding programmes of broiler chickens.
justifies the use of factor scores for prediction. These
factors are orthogonal to each other and are more Table 4. Factor scores for the prediction of carcass weight
reliable in weight estimation. The two factors selected of Arbor Acre broiler chickens intensively reared
were found to have significant (p<0.01) positive at the Poultry Unit of the College of Agriculture,
linear relationship with carcass weight (Table 5). In Lafia, Nasarawa State, North Central Nigeria.
other words, carcass weight will be expected to
Traits Factor 1 Factor 2
Live weight (kg) 1.356 - 0.404
Table 3. Explained variation associated with rotated factors
along with factor loadings and communalities for Body length (cm) - 0.213 0.043
the body measurements of Arbor Acre broiler Breast circumference (cm) 0.180 0.092
chickens intensively reared at the Poultry Unit of Thigh circumference (cm) 0.071 - 0.184
the College of Agriculture, Lafia, Nasarawa Shank length (cm) - 0.110 - 0.008
State, North Central Nigeria. Wing length (cm) - 0.588 1.285

Commu- Table 5. Multiple regression model for the prediction of

Variables Factor 1 Factor 2 carcass weight of Arbor Acre broiler chickens
nality
intensively reared at the Poultry Unit of the
Live weight (kg) 0.889 0.444 0.987
College of Agriculture, Lafia, Nasarawa State,
Body length (cm) 0.760 0.260 0.646 North Central Nigeria.
Breast circumference (cm) 0.858 0.311 0.832
Thigh circumference (cm) 0.750 0.457 0.771 Predictor Coefficient Standard t-value Probability
Shank length (cm) 0.602 0.624 0.753 error
Wing length (cm) 0.314 0.942 0.986 Factor 1 0.145 0.007 21.072 <0.01
Eigenvalue 4.63 0.62 Factor 2 0.094 0.007 13.631 <0.01
Percentage variance 77.19 10.34
Description Form Length Constant = 1.271, R2 = 84.50%, Adjusted R2 = 84.20%
966 Revista UDO Agrícola 9 (4): 963-967. 2009
 Yakubu et al. Using factor scores for predicting the carcass weight of broiler chickens using body measurements

LITERATURE CITED commercial broilers reared in deep litter system in

the derived savanna area of Nigeria. World Journal
Anonymous, 2001. Statistical Package for Social of Agricultural Sciences 4: 487-491.
Sciences. SPSS Inc., 444 Michigan Avenue,
Chicago, Illinois, USA. 60611. Pinto, L. F. B.; I. U. Dacker, C. M. R. De melo, M. C.
Ledur and L. L. Coutinho. 2006. Principal
Grice, J. 2001. Computing and evaluating factor components analysis applied to performance and
scores. Psychological Methods 6, 430-450. carcass traits in the chicken. Anim Res 55: 419-425.
Gueye, E. F.; A. Ndiaye and R. D. S. Branckaert. Rencher, A. C. 2002. Methods of Multivariate
1998. Prediction of body weight on the basis of Analysis. Second Edition, Wiley.
body measurements in mature indigenous chickens
in Senegal. Livestock Research for Rural Shahin, K. A. 2000. Sources of shared variability of
Development 10: http://www.cipav.org.co/lrrd10/3/ the carcass and non-carcass components in Pekin
sene103.htm. ducklings. Ann Zootech 49: 67-72.

Ibe, S. N. 1989. Measures of size and conformation in Shahin, K. A. 1999. Sources of shared variability in
commercial broilers. J Anim Breed Genet 106: 461- meat weight distribution and conformation in Pekin
469. ducklings. Ann Zootech 48: 143-150.

Isiguzar, E. 2003. Growth, carcass traits and meat Shahin K. A. and N. S. Hassan 2000. Sources of
quality of Bronze and White turkeys in Isparta shared variability among body shape characters at
province of Turkey. Arch Tier Dummerstorf 46: marketing age in New Zealand White and Egyptian
478-481. rabbit breeds. Ann Zootech 49: 435- 445.

Keskin, S.; I. Daskiran and A. Kor. 2007. Factor Tabachnick, B. G. and L. S. Fidell. 2001. Using
analysis scores in a multiple linear regression model Multivariate Statistics. Allyn and Bacon, Boston,
for the prediction of carcass weight in Akkeci kids. J Massachusetts, USA. p. 966.
Appl Anim Res 31: 201-204.
Teguia, A.; H. M. Ngandjou, H. Defang and J.
Kleczek, K.; E. Wilkiewicz Wawro, K. Wawro and Tchoumboue. 2008. Study of the live body weight
W. Makowski. 2007. Effect of body weights of day- and body characteristics of the African Muscovy
old Muscovy ducklings on growths and carcass Duck (Cairina moschata). Trop Anim Health Prod
traits. Arch Tierz Dummerstorf 50: 204-213. 40: 5- 10.

Manly, B. F. J. 2005. Multivariate Statistical Wolanski, N. J.; R. A. Renema, F. E. Robinson, V. L.

Methods: A Primer. 3rd Edition, Chapman and Hall. Carney and B. I. Fanchert. 2006. Relationship
between chick conformation and quality measures
Mendeş, M. 2009. Multiple linear regression models with early growth traits in males of eight selected
based on principal component scores to predict pure or commercial broiler breeder strains. Poultry
slaughter weight of broiler. Arch. Geflügelk. 73 (2): Science 85: 1490-1497.
139-144.
Wood, J. M.; D. J. Tataryn and R. L. Gorsuch. 1996.
Mendeş, M.; A. Karabayir, I. E. Ersoy and T. Savas. The effects of under and overextraction of principal
2005. The relationship among pre and post slaughter axis factor analysis with varimax rotation.
traits of American Bronze Turkey. Arch Tierz Psychological Methods 1: 354-365.
Dummerstorf 48: 283-289.
Yakubu, A. 2009. Fixing collinearity instability in the
Musa, H. H.; G. H. Chen, J. H. Cheng, B. C. Li and estimation of body weight from morpho-biometrical
D. M. Mekki. 2006. Study on carcass characteristics traits of West African Dwarf goats. Trakia Journal
of chicken breeds raised under the intensive of Sciences 7: 61-66.
condition. International Journal of Poultry Science
5: 530-533. Yang, Y.; D. M. Mekki, S. J. Lv, J. H. Yu, L. Y.
Wang, J. Y. Wang, K. Z. Xie and G. J. Dai. 2006.
Ojedapo, L. O.; O. Akinokun, T. A. Adedeji, T. B. Canonical correlation analysis of body weight, body
Olayeni, S. A. Ameen and S. R. Amao 2008. Effect measurements and carcass characteristics of Jinghai
of strain and sex on carcass characteristics of three Yellow chicken. J Anim Vet Adv 5: 980-984.
Revista UDO Agrícola 9 (4): 963-967. 2009 967