Written by Ayu Diah Sacred Kinasih-third example below is partial survey results from 3 different survey institutions, its object 1 i.e. elections jakarta, but the results are different? Which one should we believe? Why should damn?
"Survey of SMRC:P Ahok emilu DKI can still not matching the naysayers. IRMC do simulation by using the 22 respondents to prospective candidates in elections around the WNI in DKI Jakarta who have voting rights in 2017. The survey was conducted on February 24, 2017 until June 29, 2016 against 820 respondents with interview techniques. Ahok remains the most high, already a majority, 53.4%. Yusril Ihza Mahendra then 10.4% Rismaharini 5.7%, Tri, Sandiaga Uno 5.1%, 4.6%, Yusuf Mansur and another candidate under 3%. Who doesn't know as much as 9.4%. This survey used a margin of error of ± 3.9% with a confidence level of 95% "
"The institutions survey Poltracking Indonesia: Tri Rismahardini and Sandiaga Uno excelled. Simulation results from head to head couples, couples Tri Rismaharini with Sandiaga Uno superior 38.21% percent compared with a pair of Ahok Djarot just get 36.92 percent. The dilakakukan survey on the WNI in DKI Jakarta aged 17 years and above by the number of samples as much as 1,200 respondents at tangggal 6-9 september 2016. The margin of error for this survey is 2.83% at a confidence level of 95%.//
Indonesia Development Monitoring (IMD): Anies Baswedan-couples Sandiaga Uno ranked first with the acquisition of sound as much as 47.3%. Then followed the pair Agus Harimurti – Sylviana as much as 17.3%; and a pair of Basuki Tjahaja full moon (Ahok) – elected Djarot as much as 16.2%. While no vote obtained as much as 19.2%. IMD survey conducted on September 24 to september 29, 2016 with the number of Respondents from 3067 7.4 million citizens of DKI Jakarta who have voting rights in the election in 2017. the survey was conducted by using the method of multistage random sampling with 98% confidence level and a margin of error of ± 2.1%.
Now close to replicating the DKI many survey results scattered everywhere. So the various institutions are sandboxed survey results released his competing large event. Almost every day there's only institution that conducts survey results any different publications as an example of the results of the above survey. This led to many disagreements in addressing the results of the survey.
So why is the survey carried out? We always have limitations in understanding everything about the population. In the context of the science of statistics (Statistics), all these things are called parameters, namely the characteristics of the population that ─ ─ often is not known with certainty in value, especially when the unit is a member of the population is very large in number. So we need a parameter to know the parameters of the population.
For more details we will discuss further …
> > > Survey
Based on survey research techniques for SEARCHING KBBI is by giving a clear restriction upon the data, investigation and review. The survey (survey) or self-administered survey is a method of primary data collection by providing questions to individual respondents. So concludes a survey is a method of collecting information from groups representing a population.
Quantitatively survey, done by observing some – small – members of a population to gain an overview about the parameters regarding the population we want to know. A small portion of the observed population member is called a sample, its characteristics can represent a parameter of the population. Overview of population parameters were obtained based on the characteristics of the sample in principle is simply the estimated or approximate. The survey can be done by: a large number of respondents, involves asking questions to people, Using questionnaires within a relatively short time.
Survey research purposes is knowing a general overview of the characteristics of the population. Research surveys are used to gather information from a large number of opinions shaped people against a specific topic or issue. There are three main characteristics of the research survey: information is collected from a large group of people to describe certain aspects or certain characteristics, information collected through the submission of questions both written and oral from a population, information obtained from a sample, rather than the population.
The difference of the survey and the Census
In addition to the survey, another way to find out is a population census parameters required. the census is an enumeration of the population, the economy, and so on are done by the Government in a certain period of time, say ten years, done simultaneously and are thoroughly within the territory of a State to the interests of the demographics of the country concerned; divide tubers soul etc.
Examples of the activities of the census was the population census was carried out of the Central Bureau of statistics (BPS) in December 2010 (SP210). On the intended population, about SP2010 is all the residents in the territory of the unitary State of Republic Indonesia (SO) at the time the census was carried out, without exception. From about SP2010 then retrieved a number of parameters of the population (the population of Indonesia in May 2010), for example, the population according to certain characteristics: gender, age, education and region, ditamatkan
Because all members of the population observed, an idea of the population parameter that is retrieved through the census is very accurate. However, there is a price to be paid for it, which is very expensive and human resources (labor) that does little. About sp2010, for example, has spent nearly 4 trillion state money and involving about 500 thousand people the census officer. In addition, the necessity to observe all members of the population resulted in a population parameter Variant can be photographed became very limited.
> > So in General, the difference in survey and census
1. the observed sample
-Survey: only involves some samples from a population
-Census: involves the entire population
2. The cost of the
-Survey: costs relatively less
Census: the cost relatively more expensive
-Survey: relatively short
4. The Accuracy Of The Data
-Survey: not too accurate, data retrieval on the part of the population
-Census: accurate, because it involves the entire population
What causes such a survey could be wrong?
1. The Margin Of Error
Keep in mind that a survey done to the sample (sample), i.e., the portion of the population, not entirely. So, it is natural if the survey results are not exactly 100% with actual results.
Errors arising out of sampling is inevitable in any measurement of public opinion or polls. Therefore researchers must specify the particular sampling error arises from errors in the sampling. The determination of the sampling error margin was known by common error (ME). Determination of sampling error or the margin of error will have implications for the number of samples taken in representing the population of the voters in a County/city or province.
2. Non Sampling Error
Non-sampling error is an error that occurs on the outside due to the use of samples, but instead occurs when the process of the implementation of the survey. If on a survey, non sampling error occurs is very large even though sampling error/error margin is small then the set will nonetheless result to be not accurate.
Some types of non-sampling error are non-response (respondents did not respond when polled), innacurate response (respondents gave an incorrect response), selection bias (the selected respondent is not the appropriate individual/in accordance with the the purpose of the survey), dishonest interviewer (the interviewer is not honest in filling out the questionnaire), and human error (the interviewer did not deliberate errors when filling out a detailed questionnaire).
Non sampling error much due to human error in this case interviewer as the spearhead of the public opinion survey or poll. It could arise from an understanding of the material against the interviewer questions, understanding the interviewer in the procedure of determination of the sample or the ability of the interviewers in the interview process with voters (face to face interview).
To avoid the occurrence of non sampling error. There are a variety of methods to minimize non-sampling error, including the reward (gift-giving to individuals that are willing to become respondents), callbacks (contact the respondent again to validate the given response when surveyed), and trained interviewers (select and train the interviewers before conducting a survey).
3. Political Influence
Currently, the mass media have a great role in building public opinion. Therefore, many persons of a certain (political party) attempted to make use of mass media to influence public opinion. One of the tools used are the results of the survey, because the results of the survey are a result of research so that the public regarded as factual information and trusted.
Some of them are willing to pay for a survey of institutions issuing survey results which tend to favour to them to build a community mindset so that later the favour or support them. As a result, many agencies are not independent surveys (political charge) when doing a survey or poll because of the insistence on their pay. This phenomenon often occurs especially when entering an election campaign (election) as it is today.
What Should I Do?
1. Must Be Skeptical Of The Results Of The Survey
Alas, the public often not too look at things above, or may not know. In fact, the results of a survey ─ particularly those published in the media, although it could have been based on a methodology that is rocking the (shaky method) ─ has a magnitude which is very powerful in driving and influenced public opinion. In short, the necessary skepticism from the public against the results of a survey. We should not directly believe and swallow his round-round results are presented.
2. Further examination
Further study to find out whether the survey results actually is and does not have a political charge. If you want to make the survey results as one consideration in the decision. Most importantly, do not use the survey results to the bandwagon affect others or even spreading the hate just for the sake of supporting specific parties and dropped his opponent.
Results of the survey wrong can occur, either intentional or unintentional. Therefore, bijaklah to understand or make use of the results of the survey. Please use the survey results to simply expand the insight.
The author of the Ayu Diah: Sacred Kinasih, further to contact 082333878288