Rural women's radio listening behavior and program preferences in SNNPRS, the case of Sidama and Gedeo Zones

The objective of the study was to identify rural women's radio program preferences and listening behaviours. Survey was used to gather information from 200 rural women selected through multistage sampling from Sidama and Gedeo Zones. Descriptive statistic such as frequency and percentage were used to present results. Moreover, association among the different variables was tested using correlation and multiple regressions. Result from Pearson Correlation analysis indicated that there exist significant but negative association between radio listening hours and variables such as number of children, habit of listening before marriage, skill to operate radio, and education level. The result of the multiple liner regression indicated that a significant regression equation was found (F5,136)=8.679, p=0.0005) with an R2 of 0.242. Education with β-.263; p=0.001, makes the largest unique contribution to explaining the dependent, and listening habit before marriage is the next strongest unique contributor with β.242; p=0.003. Based on the results, it was recommended that health and agricultural issues should be included in programming; that more local news should be presented, that more traditional music be selected, and that program for women should be aired in the morning and evening times. the obtained


INTRODUCTION
In all the developing countries, women share significant number of the workforce in farming. They are responsible for half of the world's food production and between 60 and 80% of the food produced in most developing countries (Farm Radio Network (FRN), 2004). Rural women in developing countries are producers of about 80% of foods and are responsible for supervising about 30% of rural families (Abedi et al., 2011;Adeola and Ayoade, 2011). In Ethiopia, the role played by women in agriculture is tremendous and extremely decisive. In this country, women are considered to be the major sources of agricultural produce (Mogues et al., 2009). Insuring media access to this large sector would significantly help in the development process. With the heavy investments of international agencies, globally, there is an opportunity to make significant improvements in the quality of life and health of rural women via the mass media, especially radio. This helps women by empowering them with social, economic and political knowledge they require to better take part in the system they are part of. There are several types of information different stakeholders desire to forward to the needy women in rural areas in Ethiopia. Information demanded by rural women include variegated health issues, family planning, legal matters, political participation, human right, child upbringing, technological knowledge and many other (Oyelude and Bamigbola, 2012).
Women are highly in need of information that will lead to their political, social, and cultural empowerment (Adeola and Ayoade, 2011). In this information society, in any area of empowerment, access to information is very crucial; thus, one of the policy frameworks should be to promote equitable access for both women and men to resources, knowledge, information and services including basic needs. It is also vital to facilitate the implementation of corrective measures to address existing inequalities in access to and control over resources, as well as other empowerment opportunities (Department of Trade and Industry DTI, 2011).
With the rise of many local and community radios, the medium has proved itself to be one of the best means to access communities with development issues (Myers, 2008). Radio reaches a large number of people instantly over a wider geographical area. In areas where other media are barely accessible, the role of radio is paramount. Radio's immediacy, portability and ubiquity make it an invaluable tool in emergencies and the humanitarian aid context (Sharma, 2012;Querre, 1992). From a development perspective, women comprise a particularly important audience for radio programming. Radio"s power to make social, economic, and political information accessible to rural women remains quite remarkable (Myers, 2008).
Access to information for women is often limited due to cultural, religious and sometimes sociological factors. The use of radio, like all ICTs, is not gender neutral. There are significant differences between the way women and men use radio; and there is evidence that women have less access to radio than men. According to Myers (2008), several factors were found to negatively affect rural women listeners in Eretria, namely: men's ownership and control of radio sets, women's lower levels of education (and lack of knowledge of languages other than their mother tongue), and women's higher and more constant domestic workload which left them little time to devote to radio listening. Women's workload restricts them from concentrating on content of radio.
Audiences remain largely illusive for programmers of media producers. It is only through research that a media outlet can get a glimpse of its audiences (Webster et al., 2006). The need for audience measurement is currently largely a function of a changing media environment (Fourie, 2003;Gane, 1994;Blumler et al., 1985). According to Mytton (2007), the questions "Who is listening?" or "Who is watching?" are increasingly loosely imaginable. Knowledge of some level about the people who are watching or listening is required from the part of the media producers.
Despite the criticisms on broadcast media, radio is still a popular medium with almost 90% of the population in many parts of the world tuning in for an average of 24 h a week (Hargrave, 2000;Shingler and Wieringa, 1998). Shingler and Wieringa (1998) further point out that radio has basic qualities that make it so "enduringly popular". Chief among these qualities is radio"s ability to talk directly to the audience. Radio is less demanding compared to TV and print that require the involvement of sight. Due to this most people use radio as a background to other activities. Because radio is easier to access (in cars, the workplace, through personal headsets), it has been found to be an ideal medium for keeping people informed about breaking news stories (Fleming, 2002;Buzzard, 2002). The scholars explains that radio"s codes are purely auditory, consisting of speech, music, sounds and silence, and since, as we shall see, the ear is not the most "intelligent" of our sense organs their deployment has to be relatively simple. The media through its reach to people at large has been instrumental though not to the extent desired in supporting the movement for women emancipation by focusing neglect and marginalization of the position of the women in society (Oyelude and Bamigbola, 2012;Crisell, 1994).
Audiences should be at the heart of any plan of media programming. Audiences" lifestyle highly affects their media viewership or listening behavior. The time at which a specific demographic group of audiences tune to a media outlet helps programmers to effectively disseminate appropriate information at the appropriate time. Rural women are among the most overburdened member of the community. They are responsible for almost all activities at home including child rearing and the various chores that they are required to do. There exists a relatively commonality of activities among rural women in many parts of Ethiopia. Sharma (2012) contends that an abundance of studies has been made all over the world that shows the male dominance in media production and content. But "how gender is related to media consumption is one of the most under-theorized questions in mass communications research".
In Ethiopia, the accessibility of radio has increased in the last decade with introduction of FM radio. Many remote parts of the country are now tuned to at least two or more radio channels, some even in their local languages. This accessibility of radio motivated many rural women to develop a desire to possess radio sets. In addition of buying low cost radio sets, women now can access programs from their own mobiles. These and other factors contributed to the increase in women radio listeners in the country. It is very vital for programmers and other stakeholders to understand the preference of rural women and their listening behavior in order to produce contents in more appealing manner and in accordance with women busy working schedules. The significance of this study, therefore, lies in its attempt to identify listening behaviors and program preferences of rural women. Audience research can be used as a means of maximizing the effectiveness of public advocacy campaigns, and of improving and enhancing education and information for effective democracy and good governance (Mytton, 2007). If there is no proper study to identify the specific time rural women be free from their daily routines and get the chance to tune to radio, any message forwarded to this group of audiences would not meet its goals. This is because as women would probably be busy doing their jobs when the messages intended to them are being aired. As the development of media in Ethiopia is at its infancy, and due to limited academics in the area of media studies, research in media and radio in the country has a very long way to go. Studies that analyze audience behavior are very rare and programmers don"t seem to care as the level of competition remains insignificant. Though there is vital gender centered differences in media use across audiences, there exist very limited or no study focusing on rural women and their program preference and listening behavior in the area.
This study is, therefore, designed with an intention of investigating rural women"s program preferences and their listening behavior. It attempts to also identify specific times of a day they tune to radio. The findings of such study would be significantly important for radio programmers in the area as it provides them proper feedback on what kind of programs rural women prefer to listen to. It may also be helpful to programmers in showing them the specific time of a day women in rural areas listen to radio. This knowledge helps programmers to identify the appropriate time to produce and air contents designed to be transmitted to rural women.

MATERIALS AND METHODS
For this particular study, two zones in SNNPR, Sidama and Gedeo Zones were selected. Based on the recent Census conducted by the CSA, Sidama Zone has a total population of 2,954,136, of whom 1,491,248 are men and 1,462,888 women; with an area of 6,538.17 square kilometers, Sidama has a population density of 451.83. More than 94% live in rural area, while 5.51% are urban inhabitants; a further 0.18% are pastoralists. A total of 592,539 households were counted in this zone, which results in an average of 4.99 persons to a household. The same census report indicates that Gedeo Zone has a total population of 847,434, of whom 424,742 are men and 422,692 women; with an area of 1,210.89 square kilometers, Gedeo has a population density of 699.84. 87% live in rural areas while 12.72% are urban inhabitants (CSA, 2007). A total of 179,677 households were counted in this Zone, which results in an average of 4.72 persons to a household. In these zones, 3 districts were identified from each. Multistage technique was used to identify the zones. Selection of these districts was conducted using lottery method. These districts include Bule, Yirga Chefe and Wenago from Gedeo zone; and Dale, Hula and Aleta Wendo from Sidama zone. Gedeo and Sidama languages are dominantly spoken in the areas. Data collection settings were rural villages in these districts. Rural women in the six districts of the two zones were subjects of the study. The women who were selected Yohannes 41 did live in rural area and agriculture was their major means of living. Subjects were restricted to those who actually have access to the medium. Cross sectional descriptive survey was used in the study and quantitative data were extracted from women in the selected area by trained enumerators. The researcher identified 240 rural women (households) that responded to the questionnaire from the two zones. The decision about sample size was made based on sample size determination table developed by Bartlet et al. (2001). Bartlet and his colleagues stated that researchers may use this table if the margin of error shown in the table is appropriate for their study. Based on anticipated return rate for self-administered questionnaire of 80%, the corrected sample size came to 220. This was divided to the two zones each getting 110 respondents that fill the questionnaires. 200 papers were used for the study after some were rejected for incompleteness. Collected data were properly edited, coded and entered to the SPSS software for further analysis. Descriptive statistics such as tables and graphs were used in the analysis. Moreover, association among the different demographic and socio-economic variables was tested using different statistical tools, such as Pearson correlation and multiple regressions (Jensen, 2002). Based on the obtained results the researcher also forwarded recommendations.

RESULTS
Demographic data indicates that the age range was between 18 and 50 years, but majority were between 26-35 years of age, significant majority were at an elementary level education, 46.2% had family size of more than 5, and that 39% had 1 to 3 children. Response also indicated that 70% of respondents understand the Amharic language to some extent. Data on socioeconomic status revealed that 66% were at moderate economic status, and 61.5% claimed that their dominant income comes from cash crops.
When asked about the type of gadget respondents use to listen to radio, majority (50%) indicated that they used a normal radio set while 24.7% indicated that they listen from a tape recorder set. 97.5% stated availability of radio programs in their language in the area and 72% had a functional radio at home. While 71.5% listen to radio, 52.7% stated that they listen to radio occasionally and 37% listen to radio frequently. Respondents were asked if they had skills in manipulating radio, majority 84.2% stated that they had the skill to do so.
When respondents were asked who from family members dominantly open/shut radio sets, while 62.3% stated that any member in the household can operate and 94.5% stated that they had no restriction at all. Respondents were also asked to compare their radio listening habits before and after marriage and surprisingly enough majority 77.4% stated that they listened to radio more after marriage. Respondents were asked to indicate the specific time of a day they habitually listen to radio dominantly. While 45% said they listen to radio in the evenings, equally significant percentage, 35%, said they usually listen to radio during the morning hours. When asked about their language preferences for radio 79% state they prefer local language. Respondents were asked their demand of program content from radio and 47.9% stated they need health related issues while equally significant percent, 40.5% indicated they need the radio stations to present contents on agricultural issues. Majority"s main goal for listening to radio was to seek knowledge (58.2%) while another 22.6% listen for entertainment. Respondents were also asked to rank their preferences of radio program by genres. Accordingly, majority (76%) ranked news as their number one preference. In case of content related preference, 45.2% of respondents preferred health contents as their number one choice followed by agricultural program which was selected by 37% of respondents. 52.1% ranked local news as their first choice followed by national news with 32.2% respondents ranking it first. As far as music is concerned, 62.3% preferred traditional music while 34.9% preferred a combination of both modern and traditional music.

Results of correlation analysis
This section provides a Pearson Correlation test results of associations between radio listening hours and selected variables. Table 1 shows the correlation analysis result of some variables such as number of children, number of children sent to school, habit of listening to radio before marriage, skill to operate radio, and education level which were tested against the dependent variable, hours listening per day. The result of the Pearson correlation shows that there exist significant negative correlation between number of children and listening hours r= -0.175, significant at 0.01 level (P=036). Similarly, significant negative correlation was also found between a number of children sent to school and listening hours. Moreover, the significant negative correlation was found between listening hours and independent variables such as habit of listening before marriage [r= -0.319; significant at 0.05 level (p=000)], skill to operate radio [r= -0.239; significant at 0.05 level (p=004)], and education level [r= -0.315; significant at 0.05 level (p=000)].

Results of multiple regression analysis
A multiple liner regression was calculated to predict participants listening hours based upon education (EDUC), number of children (NCHDSC) skill to operate radio (SKLOPR) listening habit before marriage (HBFMAR) ( Table 2). Preliminary analysis was performed to ensure there was no violation of the assumptions of normality, linearity, and multicollinaity. A significant regression equation was found (F5,136)=8.679, p=0.0005) with an R 2 of 0.242. The model explained that 24.2% of variance was because of the independent variables. Education with Beta value -0.263; p=.001, makes the largest unique contribution to explain the dependent variable, and listening habit before marriage is the next strongest unique contributor with Beta value for 0.242; p=.003. Number of children and skill to operate radio scored relatively lowest Beta value (B= -186; p= 0.46 and B= 0.174; p=0.025). Children sent to school did not make significant unique contribution to the prediction (B= -0.040; p= 669).

DISCUSSION
The main objective of this study was to identify radio program preferences of rural women and their listening behaviors. Demographic and socio-economic data indicated that majority of respondents were between 26-35 years of age; 95.5% were married and that 43.5% had at an elementary level education. The fact that majority had elementary education is helpful as it implies their level of understanding radio messages. Such listeners do have a more cognitive capacity to understand messages disseminated through radio. As far as family size is concerned, 63.5% had medium level family size and 46% had more than 5 children. Lesser number of family size and that of children is believed to have the opportunity to listen for more hours as pressure from daily family and child care chores reduce. In case of perceived economic status, 66% categorized themselves as belonging to medium economic status. Radio can be among the most affordable media technology among people in this economic status. Majority indicated that they used normal radio set and that lower percentage used cell phones for radio listening indicates the lower access of mobiles to rural women in the area. There is a very high opportunity for interventionists as 97.5% stated availability of radio programs in their local languages in the area. Presence of radio in local languages is an advantage for programmers and people planning to intervene with messages. In case of frequency of listening, 52.7% of participants stated that they listen to radio occasionally while 37% of them stated they listen to radio frequently. This figure indicates that larger majority listened radio up to a desired level. The majority 84.2% stated that they had the skill to operate radio sets. The importance of this is that even when they are alone women can manage change of channels and opening and shutting operations without seeking assistance from another person. Still the presence of some women who are unable to operate radio calls for an intervention. The fact that respondents could access radio out of their home in places like shop and public transport even ensures a more profound accessibility among some. While 62.3% respondents stated that any member in the household can operate, and 94.5% stated that they had no restriction in listening radio. Such results confirm that women are active, free and independent on their viewing behaviour and preferences. The result goes against the perception that women are deprived of accessing and operating on radio due to pressure especially from their husbands. These findings also go against Myers (2008), who found that factors, such as men's ownership and control of radio sets negatively affecting rural women listeners in Eretria.
As far as demand of information from radio, majority of the respondents stated that they demand health related issues while equally significant percent (41%), stated they require radio stations to air agricultural topics. It is, therefore, helpful to suggest shift in focus and include more of programs on health and agriculture. News programs were found to be majority"s (76%) preference while the remaining favoured entertainment. Programs intended for rural women can, therefore, be programmed with news, comedy and drama being the dominant contents. Respondents' radio news type preference indicated that majority ranked local news as their first choice followed by national news with 32.2% respondents ranking it first. Therefore, including extra local and national news has the potential to attract more women. As far as music is concerned, majority (62.3%) preferred traditional music implying that programs can be acceptable if combined by traditional songs compared to other types of music.
Pearson correlation analysis for association was conducted on some socioeconomic and demographic variables against frequency of radio listening. The result indicated that there exist significant but negative association between respondents' educational level and frequency of their daily listening hours (r=-0.239), significant negative association was also found between respondents' economic status and frequency of listening (r= -0.270). Negative association between listening hours and both educational level and economic status implies that increase in education and economic status is associated with decrease in hours of listing to radio. Even though the reason why the association in both cases became negative requires further study; one possible explanation could be that educated and those with higher economic status may have other media alternatives like TV which could share time spent on radio consumption. Correlation analysis result also shows that there exists significant but negative association between frequency of radio listening and respondents' number of children, the number of children sent to school, (r= -0.174); (r= -0.209), respectively. Burden created at household level due to care for children hinders women form listening to radio. Understanding Amharic and frequency of listening to radio were also surprisingly found to be negatively associated (r= -0.174) implying that those who understand the language listen to radio lesser. Reason for this condition demands further research. There also exist significant association between frequency of listening radio and skill of operating radio (r= 0.241). This is important because it has an implication that training women on how to operate radio sets can increase frequency of listening.
The result of the multiple liner regression which was calculated to predict participants listening hours based upon education (EDUC), number of children (NCHDSC) skill to operate radio (SKLOPR) listening habit before marriage (HBFMAR) indicated that a significant regression equation was found (F5,136)=8.679, p=0.0005) with an R 2 of 0.242. The model explained 24.2% of variance because of the independent variables. The fact that the combined effect is lower calls for further study that includes more variables which could boost the predictability of listening hours. Education with β -0.263; p=0.001, makes the largest unique contribution to explaining the dependent variable implying that as more women get education, listenership would decrease. This predicts even a lower accessibility of radio in the future. Listening habit before marriage is the next strongest unique contributor with β 0.242; p=0.003. This implies that if girls are introduced to radio prior to marriage, there is a chance that they continue listening after marriage. Number of children and skill to operate radio both indicated relatively the lowest prediction score. It can be stated that reducing child care burden of women and improvement in their operating skill might increase listening hours to some extent.
Based on the results, the researcher would like to suggest the following points to be considered as a recommendation by programmers, NGOs, and other stakeholders who intend to use radio to reach women in rural contexts. As morning and evening times are found to be preferred time for radio listening for rural women, messages intended to women should be aired at these specific times. Women do have high level of demand for health and agricultural related information; therefore, programmers should develop contents on these issues and satisfy the demand of women in rural areas. Affordable radio sets should be provided for households that do not have radio on a long term installment scheme if donation is impossible. The study shows that very significant number of women in rural area use radio for information and entertainment; therefore, those who want to address women's issues need to be encouraged to use radio for any constructive intervention. Efforts should be made to help younger women develop radio listening habit before marriage as this could be continued latter in marriage life. Affordable day care services and kindergartens need to be introduced to reduce child care burdens of women. This could give women a chance to tune to radio for more hours.

CONFLICT OF INTERESTS
The author has not declared any conflict of interests.