Table 2. Experiment design for termite mortality of particleboard and weigh loss of particleboard for choice test and multiple-choice test as comparison of actual and predicted responses

Run	Variable			Response
	X₁	X₂	X₃	Y₁		Y₂		Y₃
	X₁	X₂	X₃	A	P	A	P	A	P
1	–1	–1	0	1.46	2.78	1.71	1.20	66.61	98.14
2	–1	1	0	1.17	2.41	0.53	1.15	65.21	90.81
3	1	–1	0	1.33	3.02	0.61	0.43	58.13	115.61
4	1	1	0	1.10	2.39	0.71	0.72	90.00	90.00
5	–1	0	–1	1.15	2.39	0.70	0.72	79.98	90.00
6	–1	0	1	2.03	2.07	2.65	0.55	76.49	109.68
7	1	0	–1	1.13	2.39	0.79	0.72	73.70	90.00
8	1	0	1	2.65	1.12	1.72	0.53	83.68	85.58
9	0	–1	–1	1.47	3.16	0.60	2.05	75.04	106.31
10	0	–1	1	3.08	1.72	1.23	0.79	73.70	99,69
11	0	1	–1	1.37	2.02	0.38	1.58	83.68	83.47
12	0	1	1	2.12	3.51	1.13	0.34	82.25	107.99
13	0	0	0	2.88	3.08	0.46	0.86	90.00	89.08
14	0	0	0	2.17	1.49	0.98	–0.56	90.00	100.07
15	0	0	0	2.12	1.99	0.72	0.48	90.00	85.29

X₁, adhesive concentration (%); X₂, wood vinegar concentration (%); X₃, ratio of oil palm trunk and Sengon wood sawdust; Y₁, weight loss of particle board in the no-choice test; Y₂, weight loss of particle board in the multiple-choice test; Y₃, termite mortality; A, actual value (actual value in Y₃ is data transformation with Excel in formula = DEGREES (ASIN((data actual/100)^0,5)); and P is the prediction value with response surface methodology.