Negative correlation is seen as the direct opposite of positive correlation in that it indicates that as one variable ( predictor) increases the outcome variable is seen to decrease and vice versa. It is important to note that it is also referred to as an inverse correlation. A good example that explains negative correlation is that of the quality of grades obtained in class and the frequency of TV viewing so that those students who spend a lot of time on televisions score poorly in class compared to those who spend less time in class. The graph below clearly illustrates the relationship of TV viewing and the grades scored as an explanation of negative correlation (Angst, 1999).
Zero correlation on the other hand acts as an indicator in those studies whereby no relationships exist. The variables are scattered throughout the Cartesian plane. It is denoted by a zero correlation coefficient (Angst, 1999).
There are various methods or categories of observational studies and they include: Naturalistic studies or observations where the subjects as well as the variables of study are observed in their natural states without any element of manipulation of the subjects. The major advantage with naturalistic studies is that they provide the initial steps during the carrying out of surveys and is the only hope in the cases where the ethical considerations does not allow for manipulation more especially when dealing with human subjects. The major disadvantage with this type of research is that it does not give the researcher an opportunity to control the variables through the scientific approach. The other limitation which may generate invalid results is that the subjects may get the information of what the researcher is trying to investigate and give the researcher biased answers (Kellner, 2002).
Another method commonly used in scientific as well as psychological research is the use of surveys and questionnaires. In this kind of study, questionnaires which can either be closed or open ended are administered to the randomly sampled respondents. The questions relate to the variables which might be of interest to the researcher. The major advantage of this research method is that it is relatively cheap and less laborious as the research can be conducted with the assistance of trained assistants who might not even be having knowledge on the subject being investigated by the researcher. Poorly structured questionnaires as well as poor sampling methods may affect the general results of this survey. Poor structuring of the questionnaires may be in the form of leading questions as well as extremely long questions which might be boring to the respondents and in most cases has led to incomplete answering of questions offered by the research assistant. Another setback with this method is the observed cases where the participants or respondents often try to please the researcher by giving responses that the researcher “may be in need of”.
Last but not least is that correlational method of study commonly referred to as archival where the researcher accesses the work previously done by other researchers in the form of psychological journals and publications available on request in major public and private libraries. It may involve the examination of the records of the cases of major psychological disorders in a particular state in a given year. The major advantage is that large volumes of data provides the researcher with the opportunity to study trends and at the same time, since there is no direct contact with the respondents, there is no biasness in the quality of the collected data. Since the data can be easily collected even through on line sources, the method acts as the cheapest compared to any other observational method. The major limitations with this method is the fact that the researcher has no proof on the validity of the data they are using and at the same time the variable of interest to the researcher may be missing in these records. Last but not least among the archival methods of data collection is the fact that large volumes of data complicate the whole study (Kellner, 2002).