91Ó°ÊÓ

Sample selection is a fundamental part of any experimental study, as it sets the foundation for analyzing results. In Fibiger's study, the sample selection was straightforward but effective. He selected patients from a single hospital in Copenhagen within a specific time frame, from May 1896 to April 1897. This method ensured that the population was consistent in terms of location and period, focusing purely on the effectiveness of the diphtheria serum.

The patients were divided into two groups: those admitted on even days received the new serum along with the standard treatment, while those admitted on odd days received only the standard treatment. This clear distinction based on admission days formed the basis for defining the sample groups. Such a criterion for sample selection is simple yet crucial because:

• It removes the subjective bias in patient grouping by relying on fixed and non-manipulable criteria, such as calendar days.
• It allows researchers to draw comparisons between two well-defined groups, improving the reliability of the results.

Selection bias is a significant concern in statistical analysis as it can skew the results and lead to misleading conclusions. In the context of this study, selection bias could have occurred if patients with specific characteristics, such as more severe cases, were systematically assigned to either the 'even' or 'odd' day group.

However, in Fibiger's study, the assignment was determined by the day of the week patients were admitted to the hospital, making it random with respect to their health status upon arrival. This randomness is key to minimizing selection bias. Despite the potential for variance in patient severity or other health-related factors upon admission, the use of a seemingly arbitrary system based on admission days minimizes these concerns.

Ensuring that treatment assignment is truly random helps mitigate selection bias, which strengthens the validity of the study's results and conclusions. By using a random mechanism like the day of admission, the study effectively balanced out potential unknown variables that could have favored one treatment over the other.

Experimental design is the strategy managing how the experiment is structured to answer the research question effectively. In Fibiger's study, the design involved a simple but ingenious method where patients were divided based on admission days for treatment allocation. This design was crucial as it enabled a clear comparison between the two treatments.

Several key elements highlight the strength of this experimental design:

• Control group: Patients admitted on odd days served as a control group, only receiving the standard treatment.
• Treatment group: Those admitted on even days received the new serum, alongside the standard treatment.
• Randomization: Although days might seem a predictable factor, the randomness comes from the inability to predict which patient condition will arrive on a given day, helping to randomize the groups indirectly.

The experimental design ensures that the findings are attributable to the treatment rather than external factors. It also facilitates statistical analysis, which can confirm or reject the hypothesis about the serum's effectiveness. By adopting this design, growth in scientific understanding was achieved, attributing potential differences in patient outcomes to the experimental treatment, thereby elevating the study's credibility in the scientific community.