Why might a researcher create a graphic that leads a user to overestimate the size of some effect?

The present analysis is evidence-based analysis of graphical overestimation used by the researchers and how overestimation of the size can affect the results and author perspective about the research outcomes. Researchers in past would suggest that studies use putatively softer methods that often tend to overestimate the size of findings, inconclusive findings, evidence, and limitations. The overestimation of size tends to reduce the accuracy of findings and result as report biases. Results extracted by Fanelli et al (2013) research defined the magnitude of effect sizes that affect the results of studies under the meta-analysis. the researchers concluded that overestimation of the size of some effects by the researchers could change the perspective of the reader about the facts and figures. The differences in the magnitude of sizes could lead to illusion issues. consider the example below that demonstrates the size of the study. Figure 1 below is example of overestimated size of circle when considering the results of different countries related to non-behavioral, bio-behavioral, and behavioral considerations. The countries in the analysis are United states, AS (China, Japan, South Korea, Taiwan, Singapore, and India), 15 countries of European Union, and all other countries. The size of the circle is proportional to the size of the study and in case of little discrepancy, it could affect the result differences. Some of the circles in the example are overlapping each other and make it difficult for the reader to understand the size of the study. The size of the circle is equal to ln (2/SE) where the value of zero demonstrate the perfect matching between the size and study values. The range of values for each study is between -1 to +1. Besides the size of the circle, the geographical location each value in the graph modulate the effects for the reader. The chronological order of appearances of circles and values in the study within the metanalysis induce additional effect that is inverse variance weighted relation (Fanelli & Ioannidis, 2013).

Figure 1: Magnitude of effect sizes

A little deviation from original studies could lead to higher differences. It is important to maintain the relative size of the studies. For instance, if two studies are from two different metanalyses and lead to overestimation of effects up to some extent. The size is 10 times higher as compared to the second one and due to differences in the graphical representation, it will get more weight for the larger study. The observations here would suggest that the differences under the overestimated findings. The intrinsic limitations are the evidence for inconclusive demonstration of findings. Very often researchers overestimate the effects in a specific direction and favour the outcomes in the experimental hypothesis. Many times, overestimation is caused by greater difficulty in publishing negative results. These results are more likely to be questionable for outcomes, choices, methodologies, and vibration of results from the mean outcomes depending on the selected factors.

To identify how much each primary study had overestimated the size of effect there are different methods including deviation score, expectation factor, robustness analysis, meta-analysis scaled deviation score, scaled weighting, Z-scaled deviation score, and statistical analysis. all these methods are reliable to measure the random differences and deviations in the estimated results. The negative and positive values of overestimation show directional differences in the actual size of results and fabricated overestimated results. If overestimation is minimal as measured from statistical analysis but induce significant impact on the results it could be considered as biases in the results (Aert, Wicherts, & Assen, 2019).