Is the erm “approaching significance” cheating?

The use of the statistical test for significance is an important part of scientific research. This means that the results have a significance level of 0.05 or less. The term “approaching significance” is used when findinds do not show as statistically significant, but may be close.

Using this term can be seen as cheating as it can be used by researchers to make it seem like their findings are more significant than they actually are. This can mean that the researcher is bringing their bias into it which is a problem in scientific studies as the experimenter is supposed to be objective. This may happen though as researchers do not want to admit that their research is insignificant as they may have put a lot of time and/or money into it.

Although the statistical significance could be 0.051 meaning that the results were very close to being significant so it would not be false to state that they are approaching significance, as it would be true that the results are approaching statistical significance. Although there may be no limit to how this term can be used meaning that researchers say their findings are approaching significance when they may have a statistical significance of 0.1. This means that the term can be seen as some form as cheating as it allows a researcher to make their findings seem more important than they actually are, even if they do it in the best intentions, it is still misleading.

Are applied research findings more valuable than theoretical findings?

The question may seem an easy choice at first, but can theoretical findings be as important as applied research findings?

I believe this is the case, and to understand why it must be first understood what applied research findinds and theoretical findings are. Applied research has a clear direction and a hypothesis that it is attempting to answer, this provides immediate answers to questions and is used in every type of scientific research. On the other hand, theoretical findings have no real direction or hypothesis, they are broad and can be used for much more.

Applied research findings can be used to attempt to cure mental and physical illnesses, and many other problems. This shows that they are obviously valuable as their results can have an instant effect on something and from that comes action, which can help society or perhaps just a single individual, either way it shows its value. However, it has to be considered that theoretical research serves as a precursor to applied research meaning that it is as valuable, as without the theoretical the applied research may not be carried out, as the idea has to come from somewhere. This means that applied research findings cannot be more important that theoretical findings, because if they are derived from them then without the theoretical findings the applied research would not come to be.

Is it good science to keep adding participants/manipulating data until you find an effect?

You would assume that the researcher would not influence the results the results of his data in such way, but do some researchers do this?

Researchers are supposed to remain neutral when conducting their research to ensure that they do not carry out their research with bias, such as Rosenhan did in 1973, where he already had an idea of what the results would be, thereby causing his research to have some bias and come under fire from critics, such as Spitzer (1975). It has to be asked though whether researchers can remain neutral when carrying out research as they must already have some sort of idea of how the results will turn out.

Adding participants may or may not help to show an effect in research. Adding participants may increase the correlation between the two variables and show the effect the researcher may want, but on the other hand it could also decrease the correlation and disprove the hypothesis.

Manipulating data would be frowned upon by most scientists and not deemed as acceptable. The data may be manipulated in the media, but should not be manipulated by the researcher involved.

Manipulating data and adding participants is not good science if it’s for the use of finding an effect, as it shows that the researcher is purposely interfering with the data in order to reinforce their hypothesis.

References:

Rosenhan, D. L. (1973), On being sane in insane places, Science, 19

Spitzer, R. L. (1975), On pseudoscience in science, logic in remission, and psychiatric diagnosis: A critique of Rosenhan’s “On being sane in insane places“, Journal of Abnormal Psychology, 84(5).

The effect of the Media on the public perception of research

The media has often been said to distort the public’s view on research. This can be the case but the media can be informative to the general public. Livingstone (1996) stated that on average people spend 25 hours watching television and also read newspapers,magazines and listen to the radio.

The ever-growing media can increase the public interest in areas of research, such as psychology, helping them to grow and improve. This increases funding behind the research possibly because of more government grants or because of other investors. This shows that the media can have an important positive effect on researching.

On the other hand, the media’s control of what the majority of people believe can have negative effects on research by taking the findings out of context. Nelson, Clawley and Oxley (1997) said that the frame in which information is published control and influence opinion by putting importance on certain values, facts and other considerations. This can cause a minor correlation to become a causality and can manipulate the general public’s views on certain issues.

References:

Livingstone, S. (1996) On the continuing problems of media effects research. In J. Curran
and M. Gurevitch (Eds.), Mass Media and Society. London: Edward Arnold. Second edition.

Nelson, T.E., Clawson, R.A., and Oxley, Z.M. (1997). Media framing of a civil liberties and its effect on tolerance. American Political Science Review, 91, 567-583

The relationship between causality and correlation

Causality is the relationship between a first event which leads up or causes a second event, whereas correlation is a broad term that implies a statistical relationship between two events, where one is not necessarily cause by the other but there is some sort of relationship.

It is possible for causation and correlation to both exist in a study but can causation stand without correlation or vice versa?

Correlation can stand alone with causation, Shakir (2005) noted that there is a correlation between the amount of oxygen in a room and the ability to start a fire in said room, but this is not causation because more than just oxygen is required to start the fire. This shows that correlation can exist without causation. The idea of causation without correlation is much harder to imagine because of the fact that for one event to cause another there would also have to be correlation as there would be a clear statistical relationship between the two variables. This means that causation cannot stand alone from correlation due to the fact that there would be a relationship present.

Shakir S.A.W (2005), Chapter 7. Causality and Correlation in Pharmacovigilance, Stephens’ Detection of New Adverse Drug Reactions Fifth Edition

God – The presence of a God spot in the mind.

There has often been debate about the existence of God, but there has been an increasing interest in the possibility of the God spot, a part of someone’s brain that is linked with God or belief in God. Beauregard and Paquette (2006) carried out a study with Carmelite nuns in order to see if there is a God spot.

This study was titled “Neural correlates of a mystical experience in Carmelite nuns”, this title was suitable for the design of the study as it the idea behind the study was to attempt to discover the neurological factors behind mystical experiences. The participants were 15 Carmelite nuns, whose neural activity was monitored using MRI scans, to understand if there was a visible link between neural activity and RSMEs (religious/spiritual/mystical experiences). The findings do appear to show a correlation between neural patterns and mystical experiences, but these neural patterns reflect the changes in cognition, perception and emotions, such as joy and unconditional love (Stace, 1960). This can suggest that the neural reactions are merely a self-induced euphoria created by the nuns due to their belief in God.

The Telegraph published an article based on this study called “Nuns prove God is not a figment of the mind”, this title suggests that the study proved that God isn’t imaginary, therefore meaning that there is a God. This was not the case at all, as nothing can truly be proven in science and Beauregard even said that the study “does not confirm or disconfirm the existence of God” (Beauregard & Paquette, 2006). This is not warranted from the findings of the study as the study does not prove that the neural patterns are linked to the presence of God or a God spot, they simply show that there is some sort of neural reaction.

References:

Beauregard, M., & Paquette, V. (2006). Neural correlates of a mystical experience in Carmelite nuns. Neuroscience Letters, 405(3), 186-190. doi: 10.1016/j.neulet.2006.06.060

Highfield. R. (2006). Nuns prove God is not a figment of the mind. Science Editor. Telegraph Media Group, (2011)

Stace, W.T. (1960), Mysticism and Philosophy, Macmillan, New York.

Objectivity and Subjectivity

Objectivity is the idea of not being influenced by personal opinions or feelings. This is seen as very important in science because it provides a more valid conclusion, whereas subjectivity is the opposite where feelings and opinions influence the researcher in their conclusion.

Subjectivity can sometimes be subconscious to the researcher who may be unaware of the bias they are actually implementing in their research. This is seen in Rosenhan (1973), where Rosenhan and his colleagues went into the institutions with set ideas meaning that results would show what they had already thought before they started the experiment. This can make a study less valid because their preset thoughts on what the outcome would be affects how close to the truth the results will be, because of this objectivity is more valid method.

Objectivity can be hard to properly utilise, because the researcher must have an opinion in order to want to carry out a study. What can prevent bias in the research is by attempting to find evidence against the hypothesis of the experiment. This increases the validity of the study as it means that the researcher is not just looking for evidence to back up their theory but looking for evidence against it as well.

References:

Rosenhan, D.L. (1973) On being sane in insane places Science, 179. 250-258

Survey Data Collection

There are numerous ways to collect data for a survey including online surveys, telephone, mail and others. More than one may be used in the survey, but which is the most reliable method for collection data.

Using online surveys is cheap, produces quick results and can be used for a large sample. There are problems with this though as the participant could manipulate the results by doing the survey more than once unless the survey was protected by a password or restricted one vote per IP address. This method may also provide more younger participants than older participants, meaning the results may not be valid. There are also some households that do not have the internet, which would cause a bias in results.

Using the telephone for surveys can be effective as the presence of an interviewer can provide a higher response rate. The presence of the interviewer could also have negative effects as the participants may be more comfortable talking to a female instead of a male or they may lie to gain the interviewer’s favour. The cost of this can also vary. The information gathered from this can also be difficult to use statistically.

Questionnaires can also be sent by mail. This is not costly and means there is no interviewer bias, where it can still produce large amounts of information. Despite having advantages using mail questionnaires can be a lengthy process as it may take a long time before they are returned with some not being returned. This method is also not very useful in a study where detail may be necessary as the participants will not know if they are supposed to go in depth or may not want to.

These methods do all have their advantages but also have disadvantages. Although, used together they may provide better results in terms of reliability and validity they may not provide the best results used individually.

Validity and Reliability

Validity is the idea of results accurately reflecting the real world. Reliability is the idea that results can be consistently repeated. These are the two things are generally both important in the conclusion of an experiment.

There can often be a misunderstanding with the two about how they work together, but reliability can be present without valid, such as if an experiment repeatedly gives the same results which do not reflect the real world then the methods used would be reliable but not valid. There is often a correlation between the two but they do exist separately.

Reliability can be more important in research, because if another researcher was to repeat an experiment and achieve the same results as the original researcher then it appears that they are doing something right. Despite this validity is still important in research because it shows the truth of what is being studied in comparison to the truth.

Sampling

Sampling is the selection of individuals from a population. This is used because researching the whole population would be too costly. Collecting data is also quicker when using samples than it would be to use the whole population. The data collected from samples can sometimes be generalised to the population.

When a researcher is selecting a sample the method they use to select their sample must be carefully considered. This is because if they use random sampling the sample may not be representative of the population and may not be reliable. The use of systematic sampling may also cause problems with research as the selection has a pattern throughout meaning that some of population do not have a chance of being selected resulting in the possibility of this method also providing data that is not generalisable.

When attempting to research an entire population, it is important to get a representative proportion of participants so the data is applicable to the population. When researching an specific population this may not be as crucial if the population being studied is mostly similar, such as researching people in a certain age, racial or economic range. Despite the fact it isn’t as important as it would be researching a wider range of people, it still is important to attempt to have an accurate proportion of different people to avoid bias towards bias results.