Introduction
Using experiments is a useful tool of social scientists for them to enhance general understanding of the social scientific research logic. Socially the experimental process of collecting and analyzing data involves creates an interesting sentiment of casual study style where activities are laboratory oriented or even involve natural occurrences.
This paper gives a brief summary of the topic “ethics and controlled experiments” that are associated with natural sciences. Although not commonly used, the controlled experimental technique has important logical values which are the basis for other commonly used techniques. Experimenting is the act of gathering information, analyzing, and making observations or conclusions. Social scientists indicate that all natural activities are undertaken experimentally (Babbie, 246).
Socially controlled experiments are known to work better in well user defined research topics. Such subject matters contain well-defined concepts and proposition especially those involving hypothetical cases. They involve explanatory as opposed to descriptive procedures. Most research tends to take the laboratories procedure as opposed to regular social events setting.
Components of the classical experiments
The basic pairs of components involved in both natural and socially classical experiments include the independent and dependent variables, pre and post-testing, experimental and control groups. In line with Babbie (264), natural experiments involve putting people in unexpected situations or settings so as to find out how they would react.
Whether present or not, an independent variable is a stimulant to dependent variable in an experiment. The experiment determines the effects of its presence or absence. The variables are limitless but they must be rationally defined say through observational methods.
Like the words indicate, the pretest begins by testing the subject in terms of dependence and later independence variables. Post testing re-measures the subject with relation to dependent variables and the differences that are noted are then compared to independent variables. This method works under assumptions that a second time analysis of an experiment seem to make it more clearer and it is probably taken more keenly thus more accuracy is achievable.
Unlike laboratory experiments, socially controlled experiments involve an experimental group to which a stimulant to be administered are observed. The observation also entails control groups which do not necessarily receive stimulus. Socially, control groups are an important guard against effects of an experiment and any event outside the experiment. Sometimes the social experimental design requires more than one experimental or control group.
The double blind experiment eliminates any possible prejudgment of results by the experimenter. The subject, results, experimental or control groups should not be known in this scenario.
Due to the difficulties of developing precise operational definitions or due to ambiguousness of social scientific measurements, employing a double-blind design becomes more appropriate.
Selecting objects
According to the norm of generalization, Bebbie (250) indicates that in social research, subject selection should be critical on comparability of experimental and control groups. “Ideally, the control group represents what the experimental and control groups would be like if it had not been exposed to experimental stimulus”.
To make this groups as similar as possible, the possibility sampling, randomization, matching or both matching and randomization can be used.
Probability sampling is rarely used because it can only work out well if the subject size is small enough, or assigned to groups. Randomization also depends on numbers and large subjects would work better in this case. Matching works well with very small numbers since it involves comparison of pairs and can still utilize the large subjects especially through the quota matrix system. Babbie (252). Matching can also be done on samples acquired randomly or through probability.
Variation on Experimental Designs
The pre-experimental research design is open to errors of interpretation or internal invalidity since it only anticipates the logic of truth. On the other hand, the internal validity issue in experiments may fail to accurately reflect the reality. External validity generalizes real world sophisticated experiments thus making them more accurate. First the internal validity may be compromised by historical events during the experiment.
Second factor is the aging of long-term subjects which are involved during the experiment. Other factors include the testing procedures, the effects of instrumentation or standardization of measurements, statistical regression which calls for experimental interfaces on the subject, Selection biases in which groups ought to be comparable at the beginning of the experiment, experimental mortalities mainly the subjects which drop out before completion but end up affecting comparisons and conclusions and lastly is demoralization or a feeling of deprivation among the control group. Generalizing experimental findings is jeopardized if test situations interact with experimental stimulus.
Strengths and weaknesses of the experimental method
The natural experiments are conducted outside the laboratory setting and this is their greatest advantage of controlling internal and external invalidity. Secondly is the ability to isolate experimental variables brought about by a group of experimental stimulus thus certain subjects have different characteristics. Lastly is that replication of findings strengthens their validity.
Some of the common weaknesses include jeopardizing the generality of the experiment especially when test situation interact with experimental stimulus. The experiments involve deception thus leading to different behaviors and lastly these are experiments that may intrude on personal setting of the subjects involved due to the un-usual situations and experiences involved. As illustrated by Babbie (230+), psychological damages on subject are evident and these forces the researcher to balance potential value of the research against any possible damages involved.
Conclusion
The types of variables in an experiment depend ultimately on the nature and purpose of the experiment. The importance of social research raises concern among observers with most questions concerning the behavior of participants. Good logical experiments should involve closely related experimental and control groups and they should also have quantitative methods with specific standardized measurements and observational procedures.
Work cited
Babbie, Earl. The Basics of Social Research 4th Edition, Belmont, CA: Wadsworth, 2007.