Case study 9: Design and analysis of experments on pollen grain germination and viability rates of the leguminous

Statistical modelling/Multiple regression analysis

The constant in the table of parameter estimates represents the intercept on the y-axis, namely the predicted value of percentage germination for zero sucrose concentration.

The parameter estimates for xsucros and xsucros2 give the regression coefficients. As shown earlier in the polynomial analysis both terms are highly significant.

The values in the parentheses in the equation below are the standard errors (s.e.) for the constant and the regression coefficients, respectively.

***** Regression Analysis *****
Response variate: m_pgerm
Fitted terms: Constant + xsucros + xsucros2

*** Summary of analysis ***

	d.f.	s.s.	m.s.	v.r.
Regression	2	14955	7477.6	72.78
Residual	45	4623	102.7
Total	47	19579	416.6

*** Estimates of parameters ***

	estimate	s.e.	t(45)
Constant	3.170	3.480	0.91
xsucros	4.417	0.465	9.50
xsucros2	-0.0906	0.0128	-7.09

The fitted equation can be written as: m_pgerm = 3.17 (3.48) + 4.42 (0.46) sucrose – 0.091 (0.013) sucrose².