Internal validity is the process of evaluating fundamental presumptions in scientific examinations mostly based on trials as investigational validity. Its major determinant in clinical trials is the Bradford Hill principle. In this principle, Inferences are assumed to have internal validity if a causal analysis on a relation meets the three highlighted principles. Relationship condition where Variable A and variable B must be related, Temporal Antecedence condition in, which Proper time order must be established and Lack of Alternative Explanation Condition where the relationship between variable A and variable B must not be attributable to a confounding, extraneous, variable. If a study shows an elevated amount of internal validity then we can conclude we have strong evidence of causality. If a study has low internal validity, then we must conclude we have little or no evidence of causality (Langbein & Feilnger, 2006).

Internal validity in itself is faced with various threats that compromise our confidence in saying that a relationship exists between the independent and dependent variables. A threat in this context refers to ways that internal validity of an experiment is jeopardized. A threat is also something that causes confounds in an experiments internal validity. Examples of threats facing internal validity are History, Maturation, Attrition, Testing among others. This study is majorly based on History as a threat to internal validity, which refers to any event, other than the planned treatment event, that occurs between the pretest and posttest measurement and has an influence on the dependent variable (Posavic, 2010). It is a threat for the one group design but not for the two-group design. This is because in the one group pre-posttest design, the effect of the treatment is the difference in the pre-test and posttest scores. This difference may be as a result of the treatment or to history. It is not a threat for the two-group design because the comparison is between the treatment group and the comparison group. If the history threat occurs for both groups, the difference between the two groups will not be due to the history event.

A history effect can arise when specific events occur between the first and second measurement that are in addition to the experimental and independent variable. This threat occurs as a result of events that happen to participants during the research, which affect results but are not connected to the inner Validity. For instance, in an extended study comparing relaxation to, no relaxation on headache occurrence. Those in the no relaxation condition sought out other means of reducing their headache occurrence took more pills. This threat affects research findings in many ways, such as, timing bias. This is where an overly long or short study time line can be affected by Timing Bias, as the duration of the study may affect the results. To avoid this, the researcher must be very familiar with the professional and empirical literature in order to know the generally accepted timelines associated with similar research. Two the Compliance bias where some study subjects will fully comply with their study obligations, others will not; this is "Compliance Bias." Differential compliance with study protocols may affect study results also. Three there is "Co-intervention or Parallel-intervention" bias, which is triggered by Subjects who are enrolled in other studies or who on their own engage in, experiences which affect the independent variable or dependent

Examples of situations that illustrate how history threat can affect the findings include, an 8 year study of naturally occurring heart disease incidence, such as, the expected number of heart disease cases that would "normally" be expected in study subjects is likely to be effected by history. For example, the death of close friends may produce a study subject to change his or her diet. Alternatively, encouragement by a spouse to live a much healthier lifestyle may cause study subjects to change dietary and exercise behaviors, thereby reducing heart disease risk and the number of cases. Another example is suppose an evaluator was studying the effects of a math education program, such as, the independent variable on underperforming 10th-graders. The study was to be completed over a nine-month academic year. There was a four-month teachers' strike. It is likely that being deprived of four months' math instruction will adversely affect these 10th-graders' math performance i.e., the dependent variable.

The researcher needs to address this threat so that it does not affect the findings. A number of ways can be used to combat it. One of the ways to minimize this threat is by adding a control group, which changes the design to two groups. For example, one group receives hippotherapy and the other one does not receive hippotherapy. The only difference between the groups will be intervention but they will have the same history effects. A good control group will most likely eliminate the internal validity threats to the single group design. If a researcher cannot add a control, group and needs to keep a single group research study design. He can minimize history threat to internal validity by recording and reporting the threat or add measurement dates such as a time series quasi-experimental design. The researcher can also use randomization procedures to help minimize the risk, assuring that outside events that occur in one group are also likely to occur in the other. Randomization procedure includes Random selection and random assignment. Random assignment helps a researcher create two or more groups of Participants that at the time of assignment is probabilistically similar on the average (Campbell & Stanly, 1963).

Langbein & Feilnger (2006) also suggested methods such as obtaining as much, information as possible about the procedural details of the research study, for example, where and when the study occurs and standardization of the conditions under which the research study is carried out will. A real example of how a researcher did adequately deal with this threat in an actual published research study is in the Analysis of institutional investors and firm innovation. This was a test of three competing hypotheses on the effects of institutional investors on firm innovation. A Sample of 135 public manufacturing firms traded on the New York Stock Exchange, American Exchange or NASDAQ. The data sources were Compact disclosure database, Money Market Directory Type, Moody's Bank and Finance Manual, Nelson's Directory of Investment Managers, Diaglo's PTS New Product Announcements database. Data for controls were also gathered from COMPUSTAT tapes. The dependent variable was Firm innovation, defined as the number of new products developed by the focal firm over the period 1990-1992.

The Independent variables were Proportions of ownership held by three categories of institutional owners. Data for these variables were collected in 1989. The data, for the independent variables were gathered one year prior to the dependent variables, which were then gathered annually for three years afterwards. In this case, the history threat is brought about by the fact that data is collected in different periods leading to timing bias. In conclusion, The History threat was controlled through use of control variables that accounted for change occurring between observation periods. Selection-history was also controlled, as subjects were not chosen based on an event of interest. The researchers' decision in this case to use control variables to combat the threat enabled the minimization of timing bias. This decision led to a soaring level of internal validity of the study indicating that there is a causal relationship between proportions of ownership held by three categories of institutional owners and Firm innovation (Lynch, 2010).

Feb 9, 2018 in Review Category