Adoption of Tef (Eragrostis Tef) Production Technology Packages in Northwest Ethiopia

Figures & data

Figure 1. S-shaped diffusion curve. Source: Adapted from (Rogers, 2003)

Figure 2. Diffusion innovation theories. Source: (Roger 2003)

Figure 3. Theory of reasoned action. Source: Ajzen and Fishbein (1975)

Figure 4. Location map of the study area.Source: Ethio Geo-spatial data

Table 1. Independent t-test for continuous explanatory variables

Variables	Adopter (n = 176)	Non-adopter (n = 48)	St. Err	Adoption decision	Adoption intensity
	Mean	Mean		t-test	Pearson (r)
Age	45.48	41.47	1.94	−2.05**	0.02 (ns)
Education level	1.52	0.73	0.42	−1.85**	0.28***
Labor	2.81	2.55	0.11	−2.2**	−0.09 (ns)
Tef farming experience	24.93	22.89	1.88	−1.1(ns)	−0.03 (ns)
Livestock number	4.944	4.11	0.32	−2.6***	0.08 (ns)
Total land size	1.61	1.41	0.068	−2.95***	−0.12*
Annual income	25,324.44	23,019.26	2207.26	−1.05(ns)	0.16**
Distance to input market	1.68	2.46	0.121	6.5***	−0.43***
Tef plot number	2.33	1.64	0.129	−5.35***	−0.16**
Frequency to DAs contact	2.14	0.87	0.18	−7***	0.55***
Number of oxen	1.99	1.43	0.125	−4.45***	0.21***
Perception (score)	3.39	3.11	0.078	−3.6***	−0.07 (ns)

Source: Computed from own survey data, 2020 *** p < 0.01, ** p < 0.05, * p < 0.1 and ns P > 0.1.

Table 2. Chi-square test for categorical explanatory variables

Variables	Response	Adopter		Non- adopter		Total		Adoption decision	Adoption intensity
		N	%	N	%	N	%	X^2 test	Mann Whitney Test
HH sex	Male	160	83.77	31	16.23	191	100	20.80***	−1.17 (ns)
	Female	16	48.48	17	51.52	33	100
	Total	176	78.57	48	21.43	224	100
Participation onDemonstration	Yes	49	92.45	4	7.55	53	100	7.94***	−2.96***
	No	127	74.26	44	25.74	171	100
	Total	176	78.57	48	21.43	224	100
Credit use	Yes	72	97.30	2	2.70	74	100	23.01***	−1.98**
	No	104	69.33	46	30.67	150	100
	Total	176	78.57	48	21.43	224	100
Farmers’ cooperative membership	Yes	153	95.03	8	4.97	161	100	92.11***	0.14 (ns)
	No	23	36.51	40	63.49	63	100
	Total	176	78.57	48	21.43	224	100
Farm related training	Yes	133	86.93	20	13.07	153	100	20.02***	−1.09 (ns)
	No	43	60.56	28	39.44	71	100
	Total	176	78.57	48	21.43	224	100
Seed multiplication membership	Yes	58	100.0	0	0	58	100	21.34***	−6.36***
	No	118	71.08	48	28.92	166	100
	Total	176	78.57	48	21.43	224	100
Off/non-farm activity participation	Yes	31	86.11	5	13.89	36	100	1.44(ns)	0.37***
	No	145	77.13	43	22.87	187	100
	Total	176	78.57	48	21.43	224	100
Radio ownership	Yes	92	85.98	15	14.02	107	100	6.68**	0.37 (ns)
	No	84	71.79	33	28.21	117	100
	Total	176	78.57	48	21.43	224	100

Source: Computed from own survey data, 2020 *** p < 0.01, ** p < 0.05, * p < 0.1 and ns P > 0.1.

Table 3. Estimates of double hurdle model for adoption of Tef production technology

Variables	First hurdle (probit)			Second hurdle (truncated)
	Coef.	Marginal effect	Robust Std.Err	Variables	Coef.	Robust Std. Err
HH_sex	0.295	0.045	0.342	Education	0.002	0.003
Education	0.067	0.009	0.058	Labour	0.002	0.010
Experience	0.021	0.003	0.014	Sqincome	0.0001	0.0001
Livestock N^o	−0.016	−0.002	0.09	N^o _Oxen	0.003	0.011
Extension	0.445***	0.059	0.109	Extension	0.033***	0.008
Training	0.697**	0.112	0.303	Training	0.024*	0.015
Demonstration	0.484	0.053	0.373	Demonstration	0.026*	0.015
Perception	0.505*	0.066	1.42	Seed_multiplication	0.036**	0.015
Land size	0.058	0.009	0.408	Distance_Mkt	−0.038***	0.009
Coop member	1.723***	0.377	0.28	Age	0.001	0.003
Credit use	0.454	0.054	0.429	Age_Square	−4.72e^−06	3.2e^−5
off farm	0.248	0.031	0.32	HH_sex	0.023	0.028
Radio_ownership	−0.084	−0.011	0.283	Plot number	−0.025***	0.008
_cons	−3.925**		1.111	_cons	0.472***	0.089
sigma_cons	0.075***		0.005
Number of Obs. = 224 Wald chi2 (13) = 90.64 Prob > chi2 = 0.0000 Log pseudo likelihood = 153.50539

Source: Own estimation result, 2020 *** p < 0.01, ** p < 0.05, * p < 0.1.

Table A1. Weights and methods of rating to calculate intensity of adoption

N^o	Packages	Recommendation	weights	Methods of rating
1	Seed type	Improved seed	0.3	Ratio of area covered by improved seed to total area covered by Tef
2	Sowing method	Row	0.25	Ratio of area covered by row to total area covered by Tef
3	plowing frequency	4 and above ha^−1	0.05	Ratio of average plowing frequency plot^−1 to recommendation frequency
4	Seed rate	Broadcast = 25 kg ha^−1 for red, 30 kg ha^−1for black	0.1^4	Ratio of recommendation seed rate ha^−1 to farmers’ actual application of seed ha^−1
		Row = 10 kg ha^−1
5	Fertilizer	Urea for Red: 75 kg ha^−1	0.125^5	Ratio of actual amount of urea applied ha^−1 to recommendation amount of urea ha^−1
		Urea for Black soil: 125 kg ha^−1
		NPS/NPSZnB/NPSB: 150 kg ha^−1	0.125^6	Ratio of actual application of NPS/NPSZnB/NPSB ha^−1 to recommendation amount NPS/NPSZnB/NPSB of ha^−1
6	Crop rotation	Rotate (revolve)	0.025	Ratio of actual crop rotated Tef plot to total Tef plot cultivated in the survey year
7	Weeding	Manual weeding (2X)	0.025	Ratio of average weeding frequency plot^−1 to recommendation weeding frequency plot^−1
Total			1.00

Source: Computed from woreda and Kebele agricultural experts, development group leaders, and model farmers, 2020.

Table A2. Omitted variable test using Ramsey reset for 2nd hurdle

Hypotheses	Ramsey RESET result	Decision
H0: Model has omitted variables	F (3, 207) = 2.10 Prob > F = 0.1017	Reject H0
H1: Model has no omitted variables

Source: Computed from own survey, 2020.

Table A3. Multicollinearity test using vif for 1st hurdle (probit) variables

Variables	VIF	1/VIF
Frequency to extension contact	1.13	0.88
Education	1.09	0.92
TLU	1.26	0.79
Tef farming experience	1.22	0.81
Perception score	1.05	0.95
Land size	1.47	0.68
Mean VIF	1.20

Source: Computed from own survey, 2020.

Table A4. Breusch-Pagan/Cook-Weisberg test for heteroscedasticity for 2nd hurdle

Hypotheses	Breusch-Pagan test result	Decision
H0: Constant varianceHl: No-constant variance	chi2(1) = 26.20 Prob > chi2 = 0.0000	There is no any evidence to reject H0

Source: Computed from own survey, 2020.

Table A5. Multicollinearity test using VIF for 2nd hurdle (Truncated) variables

Variable	VIF	1/VIF
Age	40.66	0.024593
Age_Square	40.66	0.024708
Number of oxen	1.35	0.739525
Extension contacts	1.32	0.759489
Labour	1.27	0.786170
Education	1.26	0.792968
Distance to cooperative	1.19	0.838751
Total tef plot	1.13	0.881943
Sqrtincome	1.07	0.932718
Mean VIF	9.97

Source: Computed from own survey, 2020

Table A6. Collinearity test for 1st hurdle (Probit) categorical variables

Variables	(1)	(2)	(3)	(4)	(5)	(6)	(7)
(1) Sex_HH	1.000
(2) Credit use	0.131	1.000
(3) participation on Training	0.014	0.070	1.000
(4) Primary coop. mem	0.356	0.397	0.235	1.000
(5) Demonstration participation	−0.005	0.189	0.311	0.114	1.000
(6) Off_farm participation	0.066	−0.070	−0.010	−0.022	0.2077	1.0000
(7) Radio ownership	0.0519	0.0568	0.3945	0.1124	0.3196	0.1814	1.0000

Source: Computed from own survey, 2020

Table A7. Collinearity test for 2nd hurdle (Truncated) categorical variables

Variables	Training	Demonstration	Seed multiplication	Sex
Training participation	1.0000
Demonstration participation	0.3115	1.0000
Seed multiplication membership	0.1179	0.1265	1.0000
Sex of HH	0.0146	−0.0057	0.1594	1.0000

Source: Computed from own survey, 2020.

Table A8. Tobit model vs. Double-Hurdle model

	Probit	Truncated reg.	Tobit	Hypotheses	Tabulated value
N^o of observation	224	176	224	H0: Tobit model is appropriateH1: DH model is appropriate	χ^2 (0.05, 13) = 22.36
Wald χ^2 (13)	90.64	229.30	-
F (13,211)	-	-	14.49
Prob > chi2	0.000	0.000	-
Prob > F	-	-	0.0000
Log Likelihood	−53.45	206.86	−13.19
Calculated value: LR = −2[lnLT- (lnLP+ lnLTR)] = > 2[(−53.45) + (206.86) − (−13.19)] = 333.20
Decision = Reject H0

Source: Own estimation result, 2020.

Table A9. Independent double hurdle vs Heckman two-stage model

Model	Hypotheses	Test value	Decision
Heckman Vs DH	H0: there is no selection biasH1: there is selection bias	[mills] lambda = 0 chi^2(1) = (0.78, 13) Prob > chi2 = 0.3782	Fail to reject H0 at 10%

Source: Own estimation result, 2020

Table A10. Recommended fertilizer rate for some Ethiopia districts

Experiment took place districts	Urea kg/hect		NPS/NPSZnB/ NPSB
	Nitosols	Vertisols	Nitosols	Vertisols
Adet (Yilmana Densa)	75	125	150	150
Bichena	75	125	150	150
Debre Tabore	75	125	150	150
Finote Selam	75	125	150	150
Este	36	-	130	-
Achefer	50	-	100	-
Dembecha	10	-	87	-
Burie	10	-	87	-
Gozamen	79	-	130	-
Dangla	10	-	87	-
Dejen	-	160	-	100
Hulet Ejunesie	-	123	-	130
Awabel	-	140	-	87
Dembia	-	96	-	87

Source: ATA and EIAR (2019.)

Table A11. Recommended management practices for tef production

Description	Recommended practices
Seed bed preparation	• Plowing 2–4 times before planting
Planting date	• Light soils: July15–24
	• In drought prone areas: July 15
	• Heavy soils: July 24-August 7
Seed rate	• 25 kg ha^−1 for Nitosols
	• 30 kg ha^−1 for Vertisols
Weeding	• 2 times hand weeding
	• 2,4-D herbicide at 0.70–1.30 L ha^−1 at early tillering stage (about 4–5 weeks after sowing)

Source: Hailu et al. (1995).

Rogers, E. M. (2003). Diffusion of innovation (5th ed.). Free Press.

 Google Scholar

Ajzen, I., & Fishbein, M. (1975). Belief, attitude, intention and behavior: An introduction to theory and research. Addison-Wesley.

 Google Scholar

ATA & EIAR. (2019). Amhara National Regional State Bureau of Agriculture 2011/12 E.C Tef (Eragrostis tef (Zucc.) Trotter) Production Manual.

 Google Scholar

Hailu, T., Ayele, M., & Assefa, K. (1995). Improved varieties of tef (Eragrostis tef) in Ethiopia: Releases of 1970–1995. Debre Zeit (Ethiopia) Debre Zeit Agricultural Research Center.

 Google Scholar