The objects of social research and scientific are different. Specifically, social science is based on people and society while scientific is research on inorganic or organic matter. The former is expected to help us build a deeper understanding of humility and comprehension of the human behaviors and our society in a systematic, multi-level and scientific method while the purpose of the latter is to discover universal rules. (Hofman in his article Prediction and explanation in social systems argued that “There is no single answer to this question because human behavior spans the gamut from highly regular to wildly unpredictable.”)That is to say, unlike nature operation rules, human behavior is hard to predict. The results of prediction was be divided by Hofman into two parts “skill world” and “luck world”. As he mentions in the article, in the skill world, under the premise of being able to measure skills, the results of prediction in principle will be successful and almost follow the perfect rule. While in luck world, success prediction is almost entirely caused by other factors which means skills cannot play a role in this world. Thus, skills and luck are totally unrelated in an extremely stylized explanation.
Due to our real-world, it exists between these two extremes, the precision of the prediction might be affected by many factors. (Take a simple example, a person who gets his salary £1000 per month, one day he found £100 on the street suddenly. From a scientific point of view, he will turn a blind eye or hand it over to the police, so his income this month is still £1000. But from the perspective of anthropology, he might put the money in his pocket then he will have £100 more this month. However, this is not the only result. If the owner finds someone takes it as his own and calls the police, the possibility this person will not only lose the £100 that he found but also face a fine punishment, which means his earnings might less than £1000 this month. ) In the above case, it can be known that human behavior is difficult to predict, and sometimes the actual results might contain the components of luck, such as the person who found the money or the owner did not call the police. We cannot hundred percent predict which situation will happen. Therefore, unlike scientific research, even we use the natural science methodology to analyze human behavior, it is still difficult to predict the result.
Besides, selection bias is the reason which causes the error exists, even social science research copies the methodologies of the natural sciences. Compared with scientific research, selection bias is more likely to occur in social research and usually can be seen during the sampling process. Since people will be divided into typical categories according to their cognitive activities, so the selected samples may not represent the entire population. This will cause the bias might occur in the research results. Nonetheless, many scholars often overlook or forget about it when they are conducting or evaluating social research studies (Geddes, 1990). In addition, approximately 250 years ago, Newton proposed his Newtonian mechanics, which lead to a revolution in the whole physics field. This revolution makes the whole universe become clocklike which can completely predictable (Almond and Genco, 1977). Generally speaking, in the field of scientific, if a phenomenon other than the running rule occurs, it will be classified as a calculation bias or other type of bias. While in social studies, most of these errors are attributed to selection bias.
One of the impressive cases of the selection bias that mentions in the course is survivorship bias, it is a kind of logical error. In the period of World War II, Professor Abraham Ward had researched the issue that how should enhance aircraft protection to reduce the chance of being shot down. After analyzing the data, he found that although the wing was the most vulnerable location on the entire aircraft, the engine was the least attacked. Therefore, Ward argued that ‘we should strengthen the protection of the engine’, but the US Navy commander had a different opinion and believed that ‘the wing should be strengthened because it is the most vulnerable position.’ Finally, they adopted Ward’s suggestion to increase engine protection and later confirmed that the decision was completely correct and this research has had a profound impact on the military field (survivorship bias, 2019).
This case tells us that the military has made a mistake with survivor bias. In general, those planes that as sampled were the aircraft that were not shot down and those planes that were shot down were not taken into consideration at all. Thus, when we have a misunderstanding of the samples that could not represent the whole group, there will be selection bias. The selection bias in social science involves human behavior and thought, so it is often ignored and hard to predict. It is one of the points that why social science research is not useful and inaccurate when the selection bias happened in the research.
Methods- quantitative and qualitative methods
The processes of social research and scientific research are the same. Both of them need to make a hypothesis at the beginning and then conduct an experimental investigation, then finally find the conclusion to prove whether the initial hypothesis is correct. Research methods can be divided into quantitative research and qualitative research. Quantitative research is using quantitative data to empirical the exam of social phenomena. This method uses statistical analysis of cases to achieve broad and valid evidence of effectiveness and credibility. For example, a questionnaire survey is one of the well-known quantitative research methods. While qualitative research is through the communication or analysis of context to generate the information or knowledge about a particular case. For instance, elite interview belongs to qualitative research. The quantitative method used quite common in the fields of scientific and social research however qualitative method rarely can be seen in scientific research.
For researchers, compared with social experiments, scientific experiments are might a little bit easier. Namely, science technology and equipment of the experiment are getting more and more advanced. Scientists can create perfect experimental conditions, control variables, remove irrelevant factors, and find out the relationship between experimental group and control group. But the quantitative researches of sociological are not like this. Social science experiments are difficult to precisely control variables like natural science. Such as when we get the data from the questionnaire survey, even if we use many complicated methods to ensure validity and reliability, we also cannot say with certainty that the realistic behavior of the research objects is consistent with the statistical results. Because the questionnaire has the characteristics of response instability and response effects (Zaller and Feldman, 1992). As for the qualitative method, it is a subjective way, to analysis the content or language to understand the actual phenomenon of the research issue, rather than calculate with numbers to conduct the research. Such as content analysis, critical discourse analysis, interview and so on. The most commonly used as the examination method in the social sciences is investigated rather than the experiment. This is also the reason why the qualitative method seldom be used in scientific research.
Since there is often a lot of debate between quality and quantification that which one is more suitable to use in the research, more and more people have been conducting qualitative and quantitative research at the same time in recent years because both of them exist advantages and disadvantages (Laurison and Friedman, 2016). Therefore, social science has more complicated research methods. The scientific features of the questionnaire survey are significant which have similar logic programs and content structures to natural science methods. Yet the questionnaire is essentially an empirical research method, so this is the method that rarely used in scientific research which means although natural science and social science have the same logic, they all have their verification means. This is the reason why social studies are difficult to imitate natural experiments because the former is more complicated. Here, I choose a case of social research which is using quantitative methods -“Mortality after the 2003 invasion of Iraq: a cross-sectional cluster sample survey” (Burnham at el, 2006) – by using social and scientific perspectives to compare and analyze the differences between the both.
In terms of research purposes, social research aims to find out how does a situation going in an area after a significant event happened, it belongs to a continuity study. Scientific research may try to understand what causes the effect and how to prevent the attack. Such as, what kind of weapon and gunfire can resist the attack or what kind of medical treatment can reduce mortality after the invasion of Iraq. The purpose of science is finding out a solution that is suitably applied to the whole society. If it is not effective as we expect which means the experimental process might be biased.
In the research method aspect, according to the article, the researchers use quantitative methods to conduct a sample survey. After a series of safety assessments and sampling, they collected the data about the deaths of the families by using a questionnaire. As the paper mention, errors and deviations are potentially exist in every investigation (Burnham at el, 2006). In this case, many factors have to take into concern such as sensitivities, security and so on, each element may affect the accuracy of the survey results. As for the scientific, these factors may not exist since human uncertainty do not contain in their samples. That is to say, if scientists want to study what kind of weapons can prevent attacks, they must focus on the type of weapon rather than writing the questionnaire to ask people which type of weapons they like to use.
After a series of investigations, the researchers will come to a conclusion. The result of social research might be a kind of phenomenon which exists in the area. For example, in the Iraq article, authors found that post-invasion mortality rates were higher than pre-invasion and gunfire was the main reason (Burnham at el, 2006). This conclusion confirms that war will increase the mortality rates however, we still cannot predict the trend of mortality rates in Iraq in the future. While the outcome of scientific research might be a kind of theory or a solution. For instance, in order to prevent the invasion, they may invent a new type of missile to help them to reduce the mortality rates. This kind of scientific study can also predict how many people can be saved with the new technology. Therefore, natural science and social science have different orientations, no matter in terms of research purposes, research methods, and conclusions. Even if social science research uses the main method of the natural sciences, it is still difficult to effectively predict future trends.
Furthermore, qualitative research is promoting by some researchers since they argue that social sciences and scientific could not be generalized after all, they have to apply the different research paradigms. Elite interviews are one of the useful qualitative methods in the research. By interviewing the main member and analysis the content what they said to collect first-hand data (Tansey, 2007). In the case of elite interviews, there are three points are different from the concept of scientific: sampling, confirmatory and definition. First of all, scientific experiments often use random sampling to ensure that each sample could represent the group. However, elite interviews are different, which logic runs against random sampling (Tansey, 2007). That is, non-probability sampling such as convenience sampling, quota sampling, purposive sampling, and snowball sampling are more suitable (Tansey, 2007). Second, compared with scientific, interview is more difficult to verify its authenticity. Because the statements of interviewees might exaggerate their words which become unreliable. In order to avoid being questioned over the accuracy, researchers have to assess in a critical way and the collected data must also carefully weigh the value to ensure its authenticity after the interview (Tansey, 2007). Finally, the elite belongs to an abstract noun, it is difficult to define which is unlike a clear definition of scientific. According to the article (Tansey, 2007), the characteristic of the elite refers to the person who is sufficient understanding of the specific field as well as the relevant structure. Even so, everyone’s definition of elite is still different. Some people believe that the person who has a higher position is elite, others think that reputation is a good tool to measure whether the person is an elite, and still others argue that only the person who is decision-making belongs to elites. But no matter which kind of standard is used to define the elite, after an effective evaluation and critical analysis the data that we have obtained will still have certain advantages to compensate and supplement other shortcomings. Hence, through the case of elite interviews, we can know that qualitative methods are more suitable to use in social research than the main methodologies of the scientific.