|
|
Results of the ‘Campaign for Dharma’ poll.
Sri Kanchi Kamakoti Peetam released a full-page advertisement in leading newspapers early in December 2004 to provide clarifications on the issues concerning the Acharya’s arrest.
Well-known media columnists, political analysts and legal experts had expressed strong opinions about how the case must be handled to ensure a fair trial. People had also expressed strong views about the role of the media in this issue. It was important to know what the ‘common man’ felt about these issues. Hence we put together a simple poll for ‘Campaign for Dharma’, and invited participation at the website www.kanchi-sathya.org
The poll featured just three statements asked participants agree or disagree
1. This case must be shifted out of Tamil Nadu if you must get justice
2. This case must be tried by a special court to accelerate verdict in a just manner
3. The media must adopt a self regulatory code of conduct and stop such trial by media
In addition participants provided simple profile information about themselves.
We received 14,264 responses over a period of 8 days. Out of these responses we ruthlessly removed all the duplicate entries and invalid e mail ids. Thus we arrived at a nett base of 9826 participants to the poll.
56% of the participants said they heard of the poll from the internet, whereas 36% said that they heard about it from the newspapers. The balance was through friends, relatives etc. There was an overwhelming agreement with all three statements.
This case must be shifted out of Tamil Nadu if you must get justice
– 93.9% Agree
This case must be tried by a special court to accelerate verdict in a just manner
- 94.5% Agree
The media must adopt a self regulatory code of conduct and stop such trial by media
- 96.4% Agree
How reliable are these results? How would an experienced market research professional rate this poll?
We approached Ms. Rama Bijapurkar, Mumbai to conduct an independent evaluation of the poll, and provided her with whatever analysis she asked for, in order to come to her conclusions
Ms Rama Bijapurkar is a well respected market strategy consultant with 27 years of experience, of which the first 18 years have been in market research and the last 9 years in market strategy consulting. She was a Director of MODE , and the Deputy Managing Director of MARG Marketing and Research Group, both organizations known for their expertise in conducting and analysing opinion polls. Rama has also been a visiting professor at IIM Ahmedabad since 1986, and she has taught the Market Research course for the MBA program there for over ten years.
We therefore thought she would be an appropriate person to take an independent look at the data and provide her analysis and observations. She has kindly agreed to do so.
We reproduce her commentary in the following pages
Commentary by Ms. Rama Bijapurkar
The results show an overwhelming “yes “ answer to all three questions, no matter how you cut the sample and analyse it. By this I mean whether you analyse the responses by state or city from which they have come, or separately by men and women, or by age group, or even by whether they are devotees or sympathizers or aware citizens, and so on.
The lowest “yes” score reached across all different categories of responses is from the Tamil Nadu geographical region, where the debate is perhaps the most intense. But even here, one would describe the answers as a thumping majority saying “yes” to all three questions - 88% yes for Q1, 89% yes for Q2 and 94% for Q 3. Even the 1305 readers of the Hindu newspaper, which has carried fairly harsh criticism of the Shankaracharya have voted 89% yes to Q1, 92% yes to Q2, and 96% yes to Q3.
This leads to the most obvious question. Before we trust results like this, can we first be assured that this sample of 9826 respondents is genuine and of good quality? This question becomes especially important since this was an internet poll, comprising responses from people who participated on their own (as opposed to being ‘caught’ and interviewed by an interviewer), by logging on to a web site. While 4438 number of responses out of a total of 14264 responses have already been eliminated as doubtful before the poll results were analysed, it is still necessary to check the genuineness and quality of the balance 9826
Before we address that question, let me explain what I mean by quality and genuineness of the sample. By ‘genuine’, I mean that they are real people and not some determined devotee sitting and stuffing the ballot box by hitting the ‘send’ button 9826 times; or even one hundred determined devotees each hitting the ‘send button’ about a thousand times. And by ‘quality’, I mean that the sample is well distributed and does not comprise only a narrow segment of, say, devotees in one state, or, say, only old people, or , say, people who have not had access to any of the media that has been critical of this issue.
The following tests have been run on the responses to test for both genuineness and quality, and the results presented below. From the results it is clear that the sample is both genuine and well spread out, and worthy of being taken seriously
Tests of genuineness of sample
- Total Responses received : 14, 264
Duplicate E mail IDS removed 546
Suspect/invalid e mail ids, (removed): 3,892
Valid sample used : 9,826
- Back Checks with the respondents via e mail to the id that they gave when voting
A random sample of 1550 respondents (equivalent to 10.8% ) the original base of 14,264 was selected. This random selection of 1550 was made using the Random Selection feature in Microsoft Excel. Thus the list of 1550 was a random computer generated list. An e mail was sent to them at the e mail address provided by them when they voted in the poll. The email asked them to confirm that they had indeed voted in the poll ( They had to choose 1 of the three statements. 1. Yes, I participated in the poll. 2. One of my family members participated using our shared e mail id 3. No. I did not participate in the poll)
Replies received : 534
Confirmed yes: 492 (92%)
Returned mail 461 (29%)
This returned mail ratio compares favourably with the number of responses eliminated due to invalid e mail ids and duplicate e mail ids (31%)
Results: I would say “satisfactory”
- Analysis of IP addresses
An IP address gives a unique identification of the location of an individual computer or a server. For example all individual computers connecting to the internet would have a unique IP address. All computers belonging to the same ‘owner’ – say a school or college or cyber café or a company - would have the same IP address.
If a large number of responses came from the same IP address there is cause to worry about ballot stuffing and hence non genuineness of the sample.
Total number of IP addresses for 9826 respondents: 6826
Number of single respondents IP addresses: 6105 : 62 % of total
“ Multiple response from the same IP” analysis
Number of mails from same IP address
Number of such IPs
% of sample accounted for collectively
163
1
1.7%
81
1
0.8%
26 – 80
0
0
21 – 25
6
1.4%
15 - 20
5
1.0%
10 TO 14
18
1.7%
6 – 9
68
6.8%
Results : Satisfactory
Tests for quality or spread of sample
- Are they all devotees, and committed partisans on this issue, or is the spectrum of attitudes on this issue represented? People who participated in the poll were asked to characterize themselves by using one of the following descriptions given below.
Participation by different descriptions
Total number of people polled
9826
Devotee of Kanchi Mutt
34.20%
Aware Citizen
29.70%
A Sympathiser
14.90%
General Observer
14.90%
Avid Reader
6.40%
Only 34% of the sample describes itself as a devotee of the Kanchi mutt, and a further 15% as sympathizers. So a half and half split of responses has been achieved. That is good because it allows an equally good sample to analyse separately, opinion of the devotee plus sympathizer versus opinion of the “lay” public i.e. those who describes themselves as avid readers, aware citizens or general observers.
- How spread out is the sample?
Type of spread
Segment and % of all responses from this segment
Comment
Geographic
T Nadu 30%
Maharashtra 12%
Karnataka 9%
AP 9%
Delhi 4%
Kerala 4%
29 other states
and UP 32%
South India 52%
Non South 48%
India 85%
Overseas 15%
People from 35 states and Union territories participated
Only half the sample was from the south.
30% of the sample was from T Nadu, but considering that it was the most negative response of all (as discussed earlier), this is not a cause for worry.
As many as 1475 responses came from overseas, from 10 countries
Cities / Towns
Bom, Del, Kol, Chennai, Hyd, 40%
Bang
Other cities 7%
Smaller towns 53%
Again points to a very well spread out responses . Chennai accounts for 20% and Kanchipuram merely 0.1% of all responses.
Gender
Men 85%
Women 15%
Very skewed in terms of men. Presumably women do not have internet access
Age
Participation by various age groups
Total number of people polled
9826
15-20 yrs
7.0 %
21-30 yrs
20.6 %
31-40 yrs
25.3 %
41-50 yrs
21.9 %
51-60 yrs
14.3 %
Above 60 yrs
11.0%
Excellent spread of all age groups
Newspaper Readership
Hindu 13%
TOI 9%
Dinamalar 2%
Dainik bhaskar 2%
20 other papers 8 %
heard through Internet not 56%
newspaper
other sources 8.5%
Well spread out sample on this count too
Analysis of the results
So far we have rigorously tested and seen from these tests that it is reasonable to assume that the sample is genuine and not localized to anyone area or interest group. Therefore let us move on to examining the results
Overall results, of total 9826 responses
Table No. 1
Overall results
Total number of people polled
9826
Yes
No
Q1. This case must be shifted out of Tamil Nadu if you must get justice.
93.9%
6.1%
Q2. This case must be tried by a special court to accelerate a verdict in a just manner.
94.5%
5.5%
Q3. The media must adopt a self-regulatory code of conduct and stop such ‘trial by media.’
96.4%
3.6%
The overall results are overwhelmingly “yes “ to all three questions.
The question to ask in such a result is “where are the “no “ responses? Are there segments of negative responses?
Where are the responses of “yes” lower than the average of the total sample?
Lowest score state: T Nadu
Lowest score City : Chennai
Lowest score newspaper readers: Hindu
All sample
T Nadu
Chennai
Hindu readers
Q1 – Y / N
Q 2 - Y/N
Q 3 – Y/N
93.9% / 6.1%
94.5%/ 5.5%
96.4% / 3.6%
88.02 / 11.98
89.21 / 10.79
93.77/ 6.23
87.25 / 12.75
88.70 /11.30
93.16 / 6.84
89.20 / 10.80
91.80 / 8.20
96.02 / 3.98
The least positive responses segments are still 80 to 90% “yes” on all counts
Therefore, let us dig a bit deeper and focus on where the “No”s could be.
What do the non devotees and non sympathizers, overall feel? Is it different from devotees and sympathizers?
1. After removing the base of 3358 devotees:
- Overall
Overall results
Total number of people polled excluding devotees
6468
Yes
No
Q1. This case must be shifted out of Tamil Nadu if you must get justice.
94.17%
5.83%
Q2. This case must be tried by a special court to accelerate a verdict in a just manner.
94.96%
5.04%
Q3. The media must adopt a self-regulatory code of conduct and stop such ‘trial by media. ’
96.46%
3.54%
2. After removing the base of 3358 devotees and 1460 sympathizers,
- Overall
Total number of people polled excluding devotees and sympathizers
5008
Yes
No
Q1. This case must be shifted out of Tamil Nadu if you must get justice.
93.25%
6.75%
Q2. This case must be tried by a special court to accelerate a verdict in a just manner.
94.23%
5.77%
Q3. The media must adopt a self-regulatory code of conduct and stop such ‘trial by media. ’
95.97%
4.03%
Poll results by different descriptions
Total number of people polled
9826
Total
Q1
Q2
Q3
Yes
No
Yes
No
Yes
No
Devotee of Kanchi Mutt
3,358
93.24%
6.76%
93.75%
6.25%
96.22%
3.78%
Aware Citizen
2,918
93.49%
6.51%
94.69%
5.31%
95.85%
4.15%
General Observer
1,466
90.79%
9.21%
91.75%
8.25%
94.88%
5.12%
A Sympathiser
1,460
97.33%
2.67%
97.47%
2.53%
98.15%
1.85%
Avid Reader
624
97.92%
2.08%
97.92%
2.08%
99.04%
0.96%
There is no change in the responses – overwhelming yes to all questions. Interestingly, all the devotees are not convinced. There is a 5% or so who say “no’ to the first two questions.
In TN, the state that has relatively the least positive responses, is there a difference between What the non devotees and non sympathizers, overall feel? Is it different from devotees and sympathizers?
Tamil Nadu state
Total number of people polled excluding devotees and sympathizers
992
Yes
No
Q1. This case must be shifted out of Tamil Nadu if you must get justice.
87.10%
12.90%
Q2. This case must be tried by a special court to accelerate a verdict in a just manner.
88.91%
11.09%
Q3. The media must adopt a self-regulatory code of conduct and stop such ‘trial by media. ’
94.25%
5.75%
“Worst case” response: Amongst non devotees and non sympathizers, who are readers of the Hindu, where the response has been least positive, relatively, what do they feel?
Hindu readers
Total number of people polled excluding devotees and sympathizers
626
Yes
No
Q1. This case must be shifted out of Tamil Nadu if you must get justice.
87.38%
12.62%
Q2. This case must be tried by a special court to accelerate a verdict in a just manner.
88.98%
11.02%
Q3. The media must adopt a self-regulatory code of conduct and stop such ‘trial by media. ’
94.25%
5.75%
The sample size of responses in this category is a healthy number of 626, allowing for meaningful analysis. Again, the responses are overwhelmingly “yes” to all questions.
The rest of the data tables show that cuts by gender or age or any other variable not discussed here all have responses in the 85 to 95% yes to all three questions.
Conclusion
In conclusion, the responses to the Internet poll have been well screened for genuineness and the 9826 responses admitted in the poll are well spread out across geography, age , and stated position on devotees / non devotees of the mutt.
Public opinion, as reflected by this poll, is overwhelmingly (90% and above on the total sample and not going below 80% in any segment of respondents) in favour of:
(1) shifting the case out of Tamil Nadu to get justice,
(2) having a special court to expedite justice and
(3) having the media adopt a self regulatory court to stop “trial by media”
Annexure
Tables representing detailed analysis of the data follow.
Table No. 1
Overall results
Total number of people polled
9826
Yes
No
Q1. This case must be shifted out of Tamil Nadu if you must get justice.
93.9%
6.1%
Q2. This case must be tried by a special court to accelerate a verdict in a just manner.
94.5%
5.5%
Q3. The media must adopt a self-regulatory code of conduct and stop such ‘trial by media. ’
96.4%
3.6%
Table No. 2
Analysis by countries
Total number of people polled
9826
Percentage of people from INDIA
85%
Percentage of people from OVERSEAS
15%
Table No. 3