Climate information: Does dissemination channels matter? Analysis of the coffee agroforestry system in the Sidama Region of Ethiopia

Figures & data

Table 1. Socio-economics characteristics of the respondents

Variable	N. Obs	Mean	Std. Dev.	Min	Max
Age (yrs)	360	46.01	11.98	20	85
Education (yrs)	360	5.72	3.69	0	16
Household Size (no)	360	6.22	2.01	2	13
Farm Experience (yrs)	360	27.95	12.30	4	70
Coffee Farm Exp. (yrs)	360	23.65	12.18	1	70
Total Farm size(ha)	360	1.08	0.97	0.125	8.35
Total annual Income (ETB)	360	10594.32	9605.25	400	51080

Table 2. Extension education advice & Cooperative membership

Attributes	Response	Percent
Provision of extension education advice on climate related issue	No	37.5
	Yes	62.5
Cooperative membership	No	43.6
	Yes	56.4

Table 3. Access to ICT-Enabled climate information dissemination channels

Dissemination channels	Access to ICT
	Yes (%)
TV	81(22.5)
FM handset radio	175(48.6)
Traditional radio	148(41.1)
Mobile phone	230(63.9)
SMS text	230(63.9)
Broacher	0 (0)*
Poster	0 (0)*
Newspaper	0 (0)*

*Don’t have access to the communication asset. Source: Owen survey 2019.

Table 4. Use of ICT-Enabled climate information dissemination channels

Dissemination channels	Users	Non- users**
	Freq. (%)	Freq. (%)
TV	62(17.2)	19 (23.5)
FM handset radio	175 (48.6)	0* (0)
Traditional radio	108 (30.0)	40 (27.0)
Mobile phone	160 (44.4)	70 (30.4)
SMS text	14(3.8)	216(93.9)
Broacher	0 (0)*	360 (100)
Poster	0 (0)*	360 (100)
Newspaper	0 (0)*	360 (100)

*Don’t have access to the communication asset, ** Calculated out of those who have access to communication asset. Source: Owen survey 2019.

Table 5. Non-ICT information dissemination channels and their usage (N = 360)

Dissemination channels	Users	Non- users
	Freq. (%)	Freq. (%)
Village leaders	314 (87.2)	46 (12.8)
Development Agents	292 (81.1)	68 (18.9)
Neighbors-friends-relatives	275 (76.4)	85 (23.6)
Indigenous knowledge forecasters	238 (66.1)	122 (33.9)
Farmers group	182 (50.5)	178 (49.4)

Table 6. Probit regression result of factors influencing access to climate information dissemination channels

Variables	Parameter estimation			Marginal effect (dy/dx)
	Coef.	Std. Err.	P>z
Age of household head	0.0263**	0.011	0.019	0.010
Sex of household head (*)	0.974**	0.491	0.048	0.367
Schooling years	−0.017	0.019	0.354	−0.007
Extension advisory service on climate change (*)	0.405***	0.146	0.006	0.157
Receive text message (*)	0.153	0.457	0.737	0.058
Ownership of TV (*)	−0.198	0.170	0.244	−0.077
Inclusion of FM handset radio App in mobile phones*	−0.009	0.142	0.946	−0.004
Membership in the Farmers’ group (*)	0.361**	0.148	0.015	0.139
Receiving phone calling*	0.666*	0.370	0.072	0.225
Total income	−1.31e-06	7.28e-06	0.857	−5.04e-07
Farming experience	−0.025**	0.010	0.018	−0.009
Constant	−1.494868	0.577	0.010
Probit regression Number of obs = 360 LR chi2(11) = 39.38 Prob > chi2 = 0.0000\ Log likelihood = −223.38364 Pseudo R2 = 0.0810				Marginal effects after probit y = Pr(ACCESSTOCLIMATEINFO) (predict) = .60267343

(*) dy/dx is for discrete change of dummy variable from 0 to 1; ***, **, and * sig at 1%, 5% and 10% level.

Table 7. Probit regression results of factors influencing the use of climate information dissemination channels

Variables	Parameter estimation			Marginal effect (dy/dx)
	Coef.	Std. Err.	P>z
Age of household head	0.001	0.006	0.873	0.000
Sex of household head (*)	0.362	0.588	0.537	0.111
Schooling years	−0.022	0.020	0.270	−0.008
Extension advisory service on climate change (*)	0.659***	0.165	0.000	0.213
Receive text message (*)	0.534	0.358	0.136	0.201
Use of TV (*)	0.563***	0.177	0.001	0.206
Use of FM handset radio (*)	0.672***	0.225	0.003	0.229
Use of traditional radio*	0.227	0.152	0.135	0.079
Use of mobile phone (*)	0.126	0.240	0.599	0.043
Neighbors as source of info (*)	−0.105	0.184	0.567	−0.037
Membership of farmers coop (*)	0.136	0.156	0.383	0.047
Constant	−1.887	0.672	0.005
Probit regression Number of obs = 360 LR chi2(12) = 65.37 Prob > chi2 = 0.0000 Log likelihood = −192.83778 Pseudo R2 = 0.1449				Marginal effects after probit y = Pr(USEDCLIINFO) (predict) = .29124524

(*) dy/dx is for discrete change of dummy variable from 0 to 1; ***, **, and * sig at 1%, 5% and 10% level.