Fundamentals of Experiment Design

Relationship between Scientific Method and Experimental Method

The Scientific Method is a systematic approach aimed at understanding the universe based on observation and logic. This approach generates knowledge through verified and tested empirical evidence.

The Experimental Method is an integral component of the scientific method that involves the collection of empirical data through direct testing. It is one of various methods used in the scientific process to validate hypotheses and generate new knowledge. Through the experimental method, theories derived from the scientific method can be tested and strengthened or, if necessary, refuted and corrected. Thus, the experimental method plays a crucial role in the formation and refinement of scientific knowledge.

Terms in Experimental Design

• Experiment: A scientific procedure designed to test a hypothesis by manipulating and controlling certain variables under specified conditions. For example, in a biological experiment, researchers might want to test the hypothesis that lighting affects plant growth. They would set up several plants under different lighting conditions and measure their growth. In an agricultural context, this could refer to an experiment to test a hypothesis like "Fertilizer X will increase rice yield by 20%". Researchers would conduct the experiment under controlled conditions, for example using the same field, the same variety of rice, and the same climatic conditions, but with and without the use of fertilizer X.
• Design: The effort or intricacies involved in creating a plan. This could include determining which variables will be manipulated, what variables will be measured, and how the data will be collected and analyzed. For example, in designing an experiment on the effects of lighting on plant growth, researchers need to decide how many plants will be used, how long the experiment will run, and what method will be used to measure growth.
• Plan: The outcome of the design process, usually contains a detailed scheme of the experiment, including variables to be studied, data collection methods, experimental procedures, and measurement tools to be used.
• Trial: The process of conducting an experiment under restricted conditions to examine the results or effects of something. It is usually used in a periodic or discontinuous context.
• Testing: The process of examining and analyzing something to determine its quality or performance. Directed towards success, not to answer how that success occurred. For example, testing seed germination, where seeds are tested to see if they can grow well under certain conditions, without needing to understand the biological processes happening behind that growth.
• Experimentation: A scientific procedure involving testing a hypothesis by manipulating certain variables and measuring the results. Directed to understand the problem through the structure of a test that is analyzed as a whole. For example, in a fertilization experiment, researchers might set up a series of experiments to understand how different types and amounts of fertilizer affect plant growth. The analysis of these experiments will involve examining the overall data to understand how these variables interact and affect the outcome.

Things needed in conducting an experiment:

• Experiment Objectives: These should be formulated clearly and usually include the hypothesis or influence to be tested. For example, "To compare the effect of two types of fertilizers, A and B, on the growth rate of corn plants."
• Treatment: This refers to the variable that will be applied or manipulated in the experiment. The treatment must be detailed and carefully designed to ensure the validity of the results. For example, in an experiment involving fertilizers, the treatment could be the "type of fertilizer" (A or B).
• Method: This refers to the procedures used to conduct the experiment and data collection. The details of the procedure should be carefully arranged and thoroughly explained. For example, "Corn plants will be planted in similarly sized pots and given fertilizer A or B. The height of the plants will be measured every week for 12 weeks."
• Experimental and Observation Units: The experimental unit refers to the individual unit that the treatment is applied to, while the observation unit is the unit that measurements are taken from. For example, in a fertilizer experiment, each pot with a corn plant might be considered one experimental unit whereas what's measured or observed, say, the height of the corn plant grown in each pot is the observation unit.
• Measurement: The data that will be collected during the experiment should be determined in advance and data recording must be done carefully. You should determine what will be measured before conducting the experiment. In a fertilizer experiment, you might measure the height of the corn plants, the number of leaves, or the total weight of the plants at the end of the experiment.
• Experimental Design: This refers to the working scheme used in the experiment, detailing how the experiment will be conducted. This can include the number of experimental units, the number of repetitions, and how the treatment will be allocated to the experimental units, how the data will be collected and analyzed. For example, you might decide to have 10 pots for each type of fertilizer and rotate the location of the pots every week to control for environmental effects.
• Justification for Design and Repetitions: The number of repetitions in an experiment should be sufficient to obtain reliable results but not too many to be inefficient. For example, if you have limited resources, you might choose to have five repetitions of each treatment instead of ten.
• Randomization: This refers to the process of randomly assigning subjects or experimental units to treatment groups to avoid bias. For example, if you decide on the position of the pots randomly, then you reduce the likelihood that environmental effects (such as sunlight or temperature) will affect the results.
• Proposed Statistical Analysis: This refers to the method of analysis that will be used after the data is collected. This method should be predetermined to ensure that the data collected fits the analysis needs and should align with the objective of the experiment. For example, if you want to compare the effects of two types of fertilizers, you might choose to perform a t-test (difference between two means test) or ANOVA (analysis of variance) if there are more than two treatments.