Because the solution to the problem is not trivial. 

one component is different, then the systems (same topic) are different. The component may be a word, a paragraph or a section of a paper

Hi, Kassu,

Conflicting results are often the consequence of doing research with inadequate sample sizes. As the sample size increases, the confidence interval tends to become smaller, meaning that the researcher can have greater confidence in the reliability of the findings.

Conflicting results also result from significant differences in the population of interest. There is no obvious reason, for example, that a rural, white, American sample would produce the same findings as an urban, black American sample.

Other apparently conflicting results can result from differences in the way variables are measured. Different instrumentation, even varying in tiny ways, can significantly effect the findings.

If the reserachers using a same concept but haven't the same comprehension the result will be different also

Actually, it is not so obvious to come to such a conclusion. Every research has its scope defined with many variables,even if one is sure that all such variables are similar, the culture plays its role, etc...

That is why it is called research work!

The main reason for that is there is no harmony between them as each one of them work alone. 

Liqaa 

Dear Kassu,

Research findings are predicted by many factors.  The research package (questions, methodology, analytical procedures) may be the same but factors such as the time of the research, the depth of the research probings, and the level of dedication to unearth the truth regarding the research questions may vary the findings in most cases. However, that is the richness associated with research- finding different perspectives of the same problem while finding different pragmatic solutions that can improve the quality of life. That is the beauty of research.

Dickson Adom

Hello dear Kassu and dear all.

Allthough all answers are plausible so far, I got the feeling the question is only partly answered in a 'by the book manner'. Of course it's a very accurate question and every one who has been enganged in an academic research has faced this fact and is also probably aware about all the above reasons explaining the variation - or even contradiction - of results on the same topic of research. But I would like to add one more (same obvious) factor: Researcher's bias!

Yes, researchers are often biased, leading the results/conclusions to a specific direction. Which of course could be confusing and cause even damages to a science.

While I do sincerely agree with most of the answers, in particular those of Zhaohao Sun, Peter Kindle, Hussin Hejase, Dickson Adom, and Stavros Karelis, I however believe that your question (Kassu Jilcha) might benefit from some further clarification. Does the 'margins of error', you are referring to, apply to quantitative or qualitative research? It is important to know also that different disciplines have different error tolerance  thresholds. We cannot possibly lump a disciple like Mathematics or Civil engineering with one like Urban planning. Problems also vary in kind. Whereas some problems have no definite answers (wicked problems), other might have definite conclusions (tame problems), And so, which exactly do you mean? [see Rittel, H.W. and Webber, M.M., 1973. Dilemmas in a general theory of planning. Policy sciences, 4(2), pp.155-169.]

 The respondents may be different, the situations may also change in change in time . research instrument may also play  a role to the change in finding

Difference in research findings are a common occurrence especially in

the academia and that is why the researcher needs to identify the

research done in their context.

Some factors that can cause the difference are:

    1. The discipline under which the researcher is carrying the

study. Different disciplines

         have different approaches which definitely determine the results.

     2. The knowledge and objectivity of the researcher

     3. The context under which the study was done.

     4.  The validity and reliability of the research methodology

     5.  The period when the study was done

The researches under lab controlled conditions should not have that startling difference but in social sciences and surveys, all the parameters including study design, researchers' way of interpretation and the research material are likely to have significant impact on the outcome. The gaps need to be identified for further research.

Different people think in different ways and use different tools and techniques for doing research.

There are lots of reasons to have different results on same topics. It ranges from the objectives to the handling procedures. The discipline, instrumentation, laboratory facilities are playing the vital role. To me the handling of samples and equipment is very important to obtain the best result. Different researchers using the same facilities can get different data due to their handling differences.

Thanks and regards

I agree with my colleagues. That the differences might be be related to:

1. Population characteristics

2. Methods used

3. The care in data collection, data analysis and selective reporting of results, 

4. Hidden confounders which may give erroneous results

5. Bias. Systematic error

etc.

In some instance it is true that differences do exist for example in measures utilization of health facility in one population can be different due to variation in need, culture, quality of care......and so on.   

Many reasons contribute to conflicting results or findings which some of the participants on this forum already mentioned above. Some reasons may be due to sample size, inclusion or exclusion criteria, or missing data issue.

The different analytic approaches as well could be another possible reason, as it was proved to influence results even when the same dataset is used

Silberzahn, R., Uhlmann, E. L., Martin, D. P., Anselmi, P., Aust, F., Awtrey, E., … Nosek, B. A. (2018). Many Analysts, One Data Set: Making Transparent How Variations in Analytic Choices Affect Results. Advances in Methods and Practices in Psychological Science, 1(3), 337–356.

https://doi.org/10.1177/2515245917747646

Sample size, Study settings, methodological approaches, and population characteristics could be major factors responsible for differences in results from same study conducted by different researchers.