Research Design Principles

Well prepared experiment often leads to good results. In this lecture, I was able to know lots of techniques and considerations before starting an experiment. Planning is very essential before the onset of the experiment to obtain the important details of the data needed, and to achieve valid and objective conclusions. This is very crucial especially when formulating the design of the experiment to have reliable results and to avoid experiment failures. It is very important to have a well-planned experimental design because poor experimental design or planning often waste time, money, effort, and creates ethical issues.

One of the most important considerations before jumping into actual experiments is the process of knowledge acquisition. First, I need to formulate questions or recognize a certain problem and think of possible answers. Formulated questions frame both the problem and solution. The next thing I should consider is the kind of experimental method to perform to prove my hypotheses. Experimental method using smallest number of trials has the maximum efficiency. It is also very important for me to progressively acquire knowledge in order to improve and lessen possible mistakes or discrepancies I might encounter during the actual experiment. Also, one of the most important concerns I should consider is analyzing the results. It is very important since reliability of the result is dependent on the kind of analysis used, and this provides the readers an easy understanding of the results.

Taking in consideration those techniques can help me, as a student, in my future research. I was able to dissect necessary things to have good experiment. It should have well-planned inquiry to confirm or deny the results of previous experiment, and undergo series of test to check the process or system of the experimental design.

In a system or process, there are factors that I need to take into consideration. These factors arethe following:

  1. Output factor corresponds to the response variables (e.g growth, photosynthetic rate)
  2. Input factor levels of treatment variable (e.g temperature, pH, salinity, etc.)
  3. Controllable factor can be set at specified levels and these basically refers to the input applied in a system in a controlled way to know their effect on the responses (https://shodhganga.inflibnet.ac.in/bitstream/10603/121791/15/15_chapter_04.pdf)
  4. Uncontrollable factors are factors that cannot be controlled but can have an effect on the subject (e.g typhoon, power interruptions, etc.)

Diagram of the process and system

There are two categories of studies I learned in statistics: Designed experiment/Comparative studies and Observational studies.

I was able to distinguished designed experiment from observational studies through the different methods and techniques applied. In designed experiment, researcher attempts to copy or either create an environment for a certain study. Treatments were applied in designed experiments in order to know the response of the subject organisms within an artificial or manipulated environment. On the otherhand, observational studies rely mainly on existing conditions of organism of study with no control on the organism’s behavior and action. Both study categories has its own strengths and weaknesses. In scientific studies, ethical issues are crucially considered, especially when the study might impose threat to the well-being of the people involved (https://www.iwh.on.ca/what-researchers-mean-by/observational-vs-experimental-studies).

Most of the studies conducted in the laboratory employs designed experimental studies to know the different responses of the subject organism towards the certain treatment. Before doing the actual experiment, it is very necessary to consider the categories needed in conducting the designed categories.

Designed Studies Categories:

  1. Preliminary/Screening

    This is usually done before the actual experiment. In my undergraduate thesis, I was able to do preliminary experiment before jumping into my actual and final experiment. This is very important because this helps me to decide better methods and techniques to apply in order to lessen the probability of having a failed experiment.

  2. Critical

    This compares responses to different treatments. Usually done to know the range of values of the treatment variables. I was able to do this

  3. Demonstrational

    This compares the new treatment to the existing standard treatment from previous studies.

Planning for an experiment or research

It is very important to have a well-planned technique in doing the designed experiment. biological tag research programs

Checklist before doing the actual experiment: 1. specific objectives 2. identification of the influential factors 3. characteristic to measure 4. specific procedure for conducting tests 5. number of repititions 6. available resources and materials

Three important to consider before proceeding to the experiment: 1. Objectives - makes the statement clear and specific 2. Methods- accurate and well-planned 3. Review of related literatures- back up study to support the hypothesis

Types of control

  1. positive- standard, ideal

  2. negative- no application, least favorable

  3. application of both

Measuring the response variable e.g growth know the biology of the organism

Simple questions to focus activities 1. what is my objective? 2. what do i want to know? 3. why do i want to know it? 4. follow up how am I going to perform task? why am i doing this tasks? direct our attention to define the role of each activity in the research study

Ingredients of Experiments 1. treatment (predictors of environment) 2. experimental units 3. randomization 4. responses

Statistics: Population and samples

Population is the set of entities of interest while sample is the subset of entities.

Types of variables

  1. Quantitative associated with scale measurement, in terms of numbers, ratio-scale Quantitative continuous (infinite number and form continuum e.g height, weight) Quantitative discrete (finite values e.g eggs of fish, number of heart beat, number of spawn)

  2. Qualitative Qualitative nominal (name or label a values) Qualitative ordinal (information on order of choice e.g scale 1-5) Qualitative categorical (cannot be measured means by feelings or emotions)

Factors used to discover the effects, various aspect of situation (e.g temperature, humidity, grazing) Levels of factors can be specific values, categories, types to be used for a given factor e.g 12%, 14%, 16% can be used interchangably.

Experimental treatment is the focus of investigationon how organism is affected by levels of treatments (e.g pH, temperature). Denote different procedures, described in terms of specific level to be used for each of the factors (e.g 5 deg cel at pH 7), has levels of factors, set of responses for the experiment.

Experiment material unit of material as a whole wherein treated by treatments e.g tanks, trees, thalli

Experimental unit organism of interest subjected to the treatments within the experiment material. where you get the measurement. maybe individual organisms, aquaria, cage, tanks, water sample of specified volume, surface of specified area, part of thalli.

Sampling unit/s some fraction of experimental unit (100 mL of water), tissue of organism

Experimental error refers to uncontrolled variation among observed responses for experimental units which have received the same treatment

Inherent variability exist in experimental material of each treatment apply variation and lack of physical

Randomness to obtain representative sample, do not include biases (e.g healthy individual)

Replication refers to the number of experimental units/materials, a treatment is applied too.

Randomization how treatments are assigned to the experimental units within the constraints of the design chosen.

Research hypotheses generate treatment design.

Hypothesis is a tentative statement that proposes a possible explanation, testable

Importance of hypothesis guide investigators to direct thought

Null hypotheses use as basis for argument but has not yet proven, no difference prediction (all equal). Alternative hypotheses statement set-up to establish like new effect compared to existing (e.g new drug is better than the existing standard drug)