What it’s all about
LinkedIn is a valuable platform for professionals to share information, network and stay informed about current trends. However, extreme caution is required when it comes to posts that use statistical data. Many of these statements are not based on solid foundations and can be misleading.
An example
One post reported that 20% of new job starters quit after 45 days. The aim of the author of the post was to raise awareness for careful onboarding and to position their own service in this context. When asked about the origin of the information, it was simply explained that the statement was based on reports from employees and several articles on this topic. No further information was provided. Our own research did not yield any results.
A general problem
Unfortunately, the type and “quality” of this answer is often given; derivations are based on unspecified principles instead of verifiable facts. The statistics are aligned with the content, not the other way round. One thing remains to be said:
- Reports by individuals are always anecdotal and subjective, which calls into question their reliability and representativeness. They can also be distorted by personal experiences, opinions and emotions, which leads to a non-representative sample.
- The size and composition of a sample play a decisive role in the validity of statistical statements. If it is small or selectively chosen, distortions automatically arise and the generalisability of the results is limited.
- A frequent point of criticism is the lack of a clear scientific methodology. In order to correctly determine the dismissal rate of new recruits in the case described, systematic data collection and analysis is required. This includes the definition of clear criteria for new entrants and the precise recording of the time of termination. Without such a methodology, the statement is difficult to verify and loses credibility.
- Statistical validity refers to the accuracy with which a study measures what it claims to measure. Without detailed information about how the data was collected and analysed, it is impossible to judge whether the claim actually reflects reality. The reliability of the data, i.e. the consistency of the results when repeatedly collected under the same conditions, is also not given if the data basis is unclear and possibly distorted.
- Statistical statements must always be considered in context. The cancellation rate of 20% within 45 days could vary greatly in different sectors, regions or company sizes. Without contextual information and comparative values, it is difficult to categorise the significance and severity of this figure. A serious statistical analysis would include comparative data from similar studies or relevant databases to put the statement into a broader context.
- Various factors can lead to resignation, including personal reasons, working conditions, economic conditions or job satisfaction. A serious statistical analysis would attempt to identify these factors and quantify their influences, rather than presenting a simple percentage figure without context.
Conclusion
Without a well-founded and methodologically sound data basis, a representative sample and a clear scientific methodology, statements are statistically questionable. A serious evaluation requires systematic and objective collection and analysis of the data, taking into account the aspects mentioned above. This is the only way to achieve reliable and meaningful results that have a right to exist in articles.
R2049 