This survey was conducted for as a part of the Media Studies course – Disinformation, Propaganda, and Rumor in the Digital Age – at Ashoka University. Smartphones, along with the introduction of cheap data plans by Reliance Jio have brought entire villages online. This has changed everything from the way they connect with others to the way they do business. The goal of this survey was to understand the role that smartphones played in the people’s consumption of news, propaganda and other information. This is necessary to understand the ways in which junk news and propaganda are spread. In addition, this survey also aims to establish how much the local population trusts information received thorough these channels and what methods, if any, they use to verify this information.
The target population for this survey are adults residing in the District of Rai, in Sonipat, Haryana. The population size was estimated to be approximately 3000 based on metrics provided in the electoral role. A sample population size of 300 was chosen in order to ensure the survey is adequately powered and representative of the target population. The surveys were conducted in four different villages within Rai: Jatheri, Rai, Asawarpur, and Badkhalsa. The surveys were conducted during the day, as in person one-on-one interviews. The sampling strategy followed was stratified random sampling. The village had clear demarcations for lower caste and upper caste neighborhoods. Both the areas were surveyed. An individual from every third structure on the road was surveyed alternating between the left and right side of the road. An attempt was made to ensure diversity in setting between homes and workplaces. There were slight variations among each of the interviewer’s methodology including the exact number of households skipped. They were asked whether they were willing to participate in a survey on news and smartphone usage. They were asked the questionnaire attached in the appendix. Attempts were made to ensure gender parity in the data collection phase by asking the women of the households to participate, and asking the males only if the female’s refused.
A total of 299 surveys were conducted over two weeks from November 24 to December 7 2019. The average age of the respondents was 32.5 years. Of the 299 respondents 209 were male and 90 were female. As far as caste was concerned the sample was predominantly upper caste with Jats numbering 103, Brahmins making up 31, and other General Castes making up 29. Reserved classes numbered 90. 226 individuals had mobile phones of which 165 were male (Refer Fig.2 in Appendix).
Non-response bias was high on the survey with the respondents largely being male upper caste members. The predominant reason for this is that a family usually has only one smartphone and it is the men of the household that use them. Another common reason for women refusing to take the survey was that they were uneducated and could not answer questions. The main reason for non-response among the lower caste community was that they could not afford a smartphone and thus felt that they could not answer the questions asked.
The most popular app (Refer Fig.1 attached in Appendix) by a large margin was WhatsApp (223), followed by YouTube (198), Facebook (194), Others/Instagram/SMS (132), TikTok (99), and Sharechat (19). Regarding the WhatsApp and Facebook groups they were a part of, the 116 of them were on friend and a similar number were on family groups. Only a total of 14 individuals reported being on political groups (including BJP, RSS, Modi specific, and Congress). The surveyed individuals seemed to predominantly use their smartphones for consuming entertainment and receive updates regarding their jobs/business. However, 94 individuals reported receiving news updates and another 20 said that they read political updates on their smartphone. Almost all of them confirmed that there was no local WhatsApp group created by the sarpanch or MLA. Major sources of election information included word of mouth (71), door to door campaigning (24), WhatsApp (35), Posters (35) and SMSs (46) (Refer Fig.3 in Appendix).
Regarding the question about Article 370, it was asked to determine if they had received a fake news about Haryana CM Khattar. The interviewer whether they (the respondent) had heard about the reading down of Article 370 and the news that Haryanvi men could now marry Kashmiri women. In case the respondent naturally brought up the (fake) news about Chief Minister Khattar mentioning the same, it was noted down. Otherwise the interviewer moved onto the next question. In this regard, only 3 individuals responded saying that Chief Minister Khattar had said so, and one of those individuals affirmed that they knew this news to be false.
This is understandable considering most of them said that they do not consume news on their phone. When asked about what sources they used to receive news updates, TV (84) and newspaper (94), and word of mouth (84) were the most common responses while only 54 reported depending on any kind of social media or their smartphone for news updates. When asked if they trust the news they received, only 73 individuals gave an unconditional ‘yes’, while the rest were skeptical to say the least. 90 individuals said that they just do not check the news that they get, most common reasons for these were that they did not care, they did not have enough time, or that they just do not trust news they get from their phone so there is no need to verify it. The survey showed a stunning level of apathy and resignation towards the biased news landscape (although the results are not significant enough to draw correlations). Some sample answers include “no, but knew that some news can be fake”, “I don’t know how to check, there is no definitive way of knowing this” and “I believe if we don’t trust the paper then what is the point of news at all. It has to be true only”. This shows how reasons for not fact checking can vary broadly among the populous. However, the broad themes aforementioned could be used to formulate a digital literacy or fact checking module for this population.
Individual case studies become important while discussing political affiliations within the surveyed population. The following examples are from survey interviews that I conducted. Rohit Kaushik, a 22 year old male Brahmin reported being part of the DU student union and a Haryana Congress supporter. Whereas his father, whom I did not survey in order to maintain diversity, was a BJP affiliate. Rohit was the one individual to correctly identify the Khattar comment as fake news. He also said that he fact checks by watching videos on YouTube. Their family WhatsApp group was the site of routine fact checking. This illustrates how friend and family groups which we routinely associate with passive consumption can be sites for critical engagement and debate. Another interesting case was of an individual named Manish, who had his shop in the affluent part of Rai. He was a 33 year old Jat and reported being on only 2 kinds of WhatsApp groups: Religious (Brahmin) and Political (BJP, RSS, local MLA). He was BJP party member and showed us a picture of himself holding a shotgun at a BJP rally. He believed that information received on his smartphone was 90% dependable. However, he interrupted the survey mid-way to ask questions about Ashoka. He hurled accusations regarding the immoral behaviour of Ashokans with regards to alcohol, drugs and cross-access in the dorms. Once the interviewer refused to answer and suggested ending the interview, the respondent returned to answering the questions albeit half-heartedly. This case reflects how the identity of interviewers is not independent of the quality of data we collect. Ashoka has a certain reputation of being liberal, moreover, members of the lower caste communities might have also felt threatened as most of the interviewers belonged to upper caste families. Although this was the only respondent who displayed visible mistrust, others might have also told the interviewer what they thought we wanted to hear or they might have hidden certain views due to fear of judgement.
Finally, the open ended nature of the questions added a layer of complexity to the survey. The answers vary broadly making extensive quantitative analysis hard. However, a limited qualitative analysis in this report has managed to uncover certain predominant themes in the answers. These could be used as the bases for formulating a more comprehensive survey with standardised answer options. Similar surveys with a larger sample and multiple choice options might reveal statistically significant results. Therefore, the quantitative and qualitative analysis in this paper must be taken as a guide for future research as the results stated herein are only generalisable to the target population of Rai, Haryana. That being said, the survey has given a comprehensive understanding of the role of smartphones in the information landscape of semi-urban Haryana, which is becoming increasingly interconnected with the larger propaganda machine in India.