In descriptive research design, there is no manipulation of variables and no attempt to establish causality. Nykiel (2007) indicated that this type of research involves studying the preferences, attitudes, practices, concerns or interests of some group of people. Bias means a slant or deviation from the true or expected. Burns & Grove (2010) says that bias in a study distorts the findings from what the results would have been without the bias.
The characteristics of just one sample population are examined in descriptive research. Boswell & Cannon (2010) says that a descriptive design describes or explains the variables being studied and provides flexibility in examining a problem from many different angles. Boswell & Cannon (2010) says that the comparative design involves no manipulation or control of independent variable, with the dependent variable being the only variable measured in two or more groups. Descriptive designs are developed to reduce the possibility and effects of bias.
Experimental research looks for a cause effect relationship between two or more variables. Nykiel (2007) noted that experimental research are set u to allow the greatest amount of control possible so that the variables selected data may be examined closely. According to Burns & Grove (2010), control is particularly important in experimental quantitative studies. This implies that the greater the researchers control over a study, the more credible the study findings. The three fundamental elements of experimental design are manipulation, control and randomization. In control, the experimenter introduces one or more controls over the experiment (Nykiel, 2007).
Moreover, Burns & Grove (2010) says that when experimental designs are used to explore causal relationships in nursing research, the nurse must be free to manipulate the variables under study. In experimental research, if the freedom to manipulate a variable for example the type, amount, or frequency of pain control measures is under someone’s else control such as a physician or the staff nurses a bias is introduced into the study. Therefore the treatment each subject receives may differ (Burns & Grove, 2010). Bias is not used in experimental design because in this type of research, studies are conducted to determine the real and the true. This implies that researchers place great value on identifying and removing sources of bias.
In quasi-experimental designs uses similar approach to true experiments in that manipulation, control and random assignment are all of importance (Newell & Burnard, 2010). However, whilst manipulation is always present in quasi-experimental approaches, control, randomization or both may be absent. Newell & Burnard (2010) noted that in definitions of quasi-experimental design manipulation and control must be present but most researchers agree that only the presence of manipulation is required for a study to count as quasi-experiments. Boswell & Cannon (2010) indicated that in quasi-experimental design, the independent variable is still manipulated, but there is no control group.
Correlation research designs are used to assess the degree of relationship between two or more variables or attributes as they occur naturally in the group being studied. Harris (2005) noted that in this method researchers do not directly manipulate variables or assign participants to groups. Nykiel (2007) says that the primary purpose of this form of research is to determine the relationship between variables. The significant difference from experimental and quasi-experiment design is that causation be established due to lack of manipulation of independent variables.
Experimental Research in Nursing
Most nursing research is not experimental although this type of design is considered to be the classic resea5rch design. This is because experimental research assumes that the relevant variables have been identified so that they can be controlled. Taylor & Kermode (2006) noted that not all of the important variables associated with nursing care have been clearly identified. Another reason why experimental research is not used in nursing is because there are many social variables of importance to health and nursing outcomes that can not be manipulated (Taylor & Kermode, 2006). In addition, Taylor & Kermode (2006) says that it is difficult to carry out experimental research in the clinical setting where random assignment to groups and standardization of research procedures may be impossible to achieve. For those reasons, the application of experimental designs to nursing research has been limited.