1. Your instructor has reviewed several strategies to conduct and report on scientific research.  Discuss the procedures in the positivistic, scientific method and the components of this research paper. List two reasons why you agree or disagree with this worldview and how you might utilize it for your future research agenda.

There are two worldviews in research. One is the positivistic worldview.  This view is a quantitative scientific method that in which reality only exists when proven with numbers or numerical data.  Quantitative researchers are seen as having the positivistic view.  In this view a person searches for a universal law of the “truth” as stated by Bailey.  This means they want to answer a specific research question by showing statistical evidence that the data may be addressed in a particular way.  They begin with a specific hypothesis that addresses one defined issue.  The research designs in this method are predetermined and structured and do not change in the course of study.  The data that are gathered are quantifiable and statistical using counts and measures. Subject samples tend to be large and the researcher has circumscribed contact with the subjects on short term basis as stated in Bailey.  Also used in this positivistic method are techniques that include experiments and quasi-experiments, structured surveys, manipulation, control, and statistical analysis of data.  The data analysis occurs at the conclusion of data collection. So we can see that at the end of the study the researcher is able to produce statistical evidence to prove a point or their research question. In doing a positivistic, quantitative method the researcher remains aware of manipulation, doing something to the subjects in the study, control, which is the experimenter’s ability to control or eliminate interfering and irrelevant influences, and randomization, which is used to reduce the risk of systematic bias from creeping into the study.  I agree with this worldview.  First of all, as in my study, I used numbers, the levels of cholesterol in the body when suicide is committed, to prove my hypothesis.  In this way I could use statistical analysis to show the truth in my conclusion. Second, I had to make sure that I used control in my study.  In using control I was able to keep certain subjects out of my study so that they would not confound or interfere with the results of my experiment.

2. We have stated that the researcher must remember the equation [dependent variables = independent variables]. Discuss the differences between independent and dependent variables and the influence one exerts upon the other.  How was this theorum influential in your study or project?

The independent variable is the variable that is manipulated to see what kind of effect it takes on the dependent variable.  This variable is also sometimes called the experimental variable or treatment variable.  The dependent variable is the variable that determines the effectiveness of the independent variable to manipulate or treat it and is also the item that is observed and measured at the beginning and the end of the study.  In my study I was researching how lowering cholesterol levels manipulates suicide.  I was able to find that my lowering cholesterol (independent variable) suicide (dependent variable) became more violent and aggressive.  I was able to easily see how my independent variable strongly manipulated the dependent variable. 

3. Create a list of the independent variables you identified in your study (Hint: these
are important client, institutional, environmental, or patient characteristics).  Give one example of a variable you could change or 'fix' in a departmental, governmental, or organizational policy.

There were many variables that could be seen as independent in my study although I was able to narrow it down to the level of cholesterol along with the depletion of serotonin.  Some other variable, besides cholesterol and serotonin, are age, dietary habits, previous detection of a depressive disorder, abuse of drugs or alcohol, and the use of cholesterol lowering drugs.  A variable that I could change is that of people using cholesterol lowering drugs without any other type of medicine along with it.  I feel that if a person’s psychological health was monitored while on cholesterol lowering therapy we may be able to change the impulsive behavior by placing them on a type of serotonin enhancer to prevent the risk of impulsive, aggressive acts. 

4. When creating a study, one must address the operational definitions for individual studies.  Give 3 examples of operational definitions you encountered in your project.  How does this process help or hinder the researcher?

In defining operational terms in a study the researcher is trying to achieve clear scientific communication so that studies can be accurately replicated.  The goal is to provide readers with information that will allow them to know what to do in order to experience that which is being defined.  Operationalizing terms allows the researcher to remove words that do not describe observable phenomena from the definition so that their meaning for a particular study can not be misinterpreted.  Three examples of operation definitions in my project were 1- parasuicide, 2- serotonin, and 3- violent/ non violent acts of suicide.  All of these definitions were needed because the words used did not describe actual phenomena that the reader would be aware of.  The words may seem familiar and a reader may be able to guess what the words mean but by defining them I allowed my researcher to me expressed accurately to all readers in order to make it able to be replicated and achieve clear scientific communication.

5. Define the different scales of measurement (i.e. Ordinal, Nominal, Interval, Ratio). In each of these scales of measure, how would the researcher decide on which statistical analysis to use?  How did you decide what methodology to use (theoretically, you told me in METHODS chapter what you decided to do with all of the datum).

Ordinal – data are numbers that still are discrete but are ordered; however the intervals are not known and cannot assumed to be equal.  In this type of data the researcher may choose to use Pearson’s chi-square, Spearman rho, Wilcoxon rank sum, Mann-Whitney, Kruskal-Wallis, or Kendall’s tau. 

Nominal – data are numbers applied to nonnumerical variables. In this data no variable or individual can be in more than one group.  Also there is no ordered relationship between categories so that one category cannot come before or after another category.  The data is also discrete and there is no limit to the number of categories that can be used. In this type of data the researcher may choose to use Pearson’s chi-square, Fisher’s exact, or Goodman and Kruskal’s tau b.

Interval – data that are in a logical sequence and the intervals are equal amounts and represent actual amounts.  This type of data is listed as continuous.  With this type of data the researcher may choose to use t tests, ANOVA (Analysis of Variance), or ANCOVA (Analysis of covariance).

Ratio – data are continuous with equal intervals between them.  There is also a zero point in this type of data. With this data the researcher may choose to use t tests, ANOVA (Analysis of variance), and Pearson product moment correlation.

--In each type of scale of measure the researcher must remain aware of what type of data that they are using.  Then they can be able to move further to see what they want to do with those data, such as show relationships or differences and between how many variables.  After the researcher decides what they are trying to do with the data they can then examine the different types of tests that answer the question they want answered and also use the data that they have.  In this way the researcher will be able to decide what kind of statistical analysis to use. 

My data were ratio data because the levels of cholesterol had equal intervals between them and that I was able to put the data on a scale containing a zero point.  My data used were parametric because they were representative of the entire population and because I was able to match patients in each group according to similar patients in other group. I wanted to be able to show a relationship between two variables, level of suicide and level of cholesterol, by doing this study. For these statistics, inferential, I used the single sample t test to compare the mean of each group to the rough population mean in order to see if the results were significant.  Also for my inferential statistics I used analysis of variance, or ANOVA, which helped me to compare the mean scores of all three groups at one time. For my descriptive statistics I simply compared the mean results to see the significance.  I decided to use this type of methodology based on what Bailey says in the book.  By looking at my data I was able to realize what type of data they were and which route to take in my scale of measurement.

6. Discuss the types of reliability. Why does a researcher in health care consider reliability an important component of their study?

Bailey states that if a study is reliable (has reliability) it is able to produce similar findings if it is repeated.  Generally reliability is concerned with replicability. Types of reliability are:

External Reliability – has to do with researchers addressing the researchers’ status position, informant choices, social situations and conditions, analytic constructs and premises, and data collection and analysis techniques.  All of these needs to be addressed in order to make sure that the study is externally reliable, or in other words that external sources and situations did not influence the results of the study in a way that would be statistically significant.

Internal Reliability – has to with multiple observers agreeing.  As Bailey says the important issue with internal reliability is that of interrater or interobserver reliability which is the extent to which meanings held by multiple observers are sufficiently congruent that they describe phenomena in the same way and arrive at the same conclusions about them.  This is to say that if two sets of researchers are doing the same study at two different sites they would still come up with the same results or conclusions. Some major threats to internal reliability are low inference descriptors, multiple researchers, participant researchers, peer examination, and mechanically recorded data. 

Some other things that can influence the reliability in a study are: 1. Subject Fatigue – subjects may be asked to perform many tasks or long tests and become tired toward the end, giving unreliable results.  2. Subject Motivation – subjects are not interested in the study and therefore have a lack of motivation to perform their best for the study.  3. Subject learning – when a study includes repeated tests the subjects will most likely begin to learn those tests causing the results from the tests to be unreliable.  4.  Subject Ability – the subjects’ abilities to respond to certain questions of tasks will be based on skill level and vary greatly.  Responses can even vary for the same subject if they decide to create a response to look better for the researchers.  5. Tester Skill – if a tester does not administer certain tests in the same manner each time the responses could become unreliable.  6. Different Testers – Testers administer tests with different levels of enthusiasm, voice, personality, etc. and therefore may cause subjects to achieve different results based on different testers.  7. Test Environment – changes in the environment, such as distractions or interruptions, can cause subjects to respond differently to tests.

A researcher in health care needs to consider reliability as an important component of their study because future researchers may use the findings from earlier studies in their study to help them come to a conclusion about the same research problem.  In order to be able to come to the most accurate conclusion possible about a research problem researchers need to be able to count on other studies to be reliable in order to further research from those earlier studies to prove their hypothesis.

7. Define validity. Examine internal and external validities and list attributes or problems associated with validity issues (think from the perspective of a potential patient or an informed peer reviewer of your study).

Validity is concerned with the accuracy of scientific findings.  In this way Bailey says that a study is only valid if investigators are truly addressing the constructs that they set out to study and measure.

Internal Validity –    1. – History – with this issue there is a study in which a measurement of an independent variable is occurring before and after some sort of manipulation or treatment and internal validity can be compromised as stated by Bailey.  A problem that can occur is when something happens to the subjects or the environment that was not planned in the study design, such as the subject becoming injured.  2. -  Maturation – this refers to time rather than to events that happen between the pre-test and post-test such as to the subject’s growth, development, or changes that occur naturally over time.  There can be problems with this type of validity in cases where the subjects are children and are more vulnerable to maturation problems.  3.  – Testing – this validity has to do with the fact that is the same test is being used several times throughout the study it is possible that subjects will experience a practice effect and score higher simply because they have been able to practice the test.  Also, in the same sense, Bailey says that if the test has to do with demonstrating strength or endurance it is possible that by performing the test the subject’s level of performance will increase.  4. – Subject Selection – subjects who volunteer for a study are likely to be different from those who are selected to participate. This can be seen as not being such a major issue because Bailey says there is some form of volunteerism for all subjects because of the consent process.  5. – Subject Mortality/Attrition – this has to do with the difficulty of trying to prevent attrition due to such things as illness or death.  Bailey states that it is important to recruit extra subjects to cover these losses in a study.  6. – Instrumentation – this has to do with internal validity issues being based on the condition of instruments themselves.  Bailey says this may have to do with an instrument being worn out or working badly or from a survey coming across as having some type of bias in it such as racial bias.

External Validities – these validities have to with the researchers trying to make their experiments controlled in order to produce sound research designs.  This can make their experiment less likely to generalize situations outside the research setting. 1. - Hawthorne Effect - With this threat subjects perform better on tests because they are receiving special attention.  This could be something that they normally have problems doing in the real world but because they are receiving attention from the researchers they try to excel in what they are doing. 2. – Replication – researchers have to report there methods section in exact detail in order for others to be able to replicate their study. If a researcher can not be assured that the same methods have been followed in replicated studies then findings cannot be generalized confidently and external validity is threatened.  3. – Generalizability - this refers to the extent that the results found in the study will also be found in the population.  Bailey says that random selection of the sample from the population is the only good way to hope for generalization.  4. – Multi-treatments -  Bailey states that in a study where a subject is given more than one treatment as the independent variable the  results cannot be generalized to other settings where only one treatment was used.  Studies that use a single treatment or independent variable are more likely to achieve external validity.  5. – Researcher Effect – subjects may react differently to the study based on their relationship with the researcher, either positive or negative.  This is even more important Bailey says if there is an interpersonal relationship taking place.

8. Discuss the characteristics of a quantitative research design. Name and discuss at least two designs from this worldview or viewpoint.  Why would you decide to use this worldview or research methodology (instead of qualitative)?

A quantitative research design is predetermined and structured and does not change throughout the study.  They are formal and specific according to a defined model and are used as a detailed plan of operation.

Two quantitative research designs are:

1. – True Experimental Designs – in this design manipulation, control, and randomization are all required.  The result is the classis cause-and-effect relationship to allow the researcher to say that there was a good chance that manipulation of the independent variable caused change in the dependent variable. The type of design also allows the researcher to compare many different types of treatments to see which one is most effective. 

2. – Non-experimental Design – this design involves no manipulation of an independent variable. Control and randomization are not possible with this design.  In this design the researcher is attempting to study a variable that cannot be manipulated because it is fixed or a variable that cannot be changed because it has already happened or a situation in which the researcher wants to compare two or more existing variables to see if there is a relationship between them.  This type of researcher is referred to as being done after the fact.  This is also called descriptive research because the researcher is simply describing the relationship without manipulating it. 

I would decide to use this type of methodology if I wanted to prove a specific research question that I had in regards to a dependent and independent set of variables.  I would be able to use a specific model to carry out my research and would be able to know that the design I was using would not change throughout the study.

9. Discuss the characteristics of a qualitative research design. Name and discuss at least two designs from this worldview or viewpoint. Why would you decide to use this worldview or research methodology (instead of quantitative)? \

A qualitative research design is general in nature rather than confined.  The design evolves through the study and remains flexible to allow for any changes to occur.  These designs are used as a “hunch” as to how to proceed. 

Two qualitative designs are:

1.  The case method research design – this design is used when the researcher wants to learn from individual clients, understand certain issues and problems in clinical settings, and also look over policies.  This type of design is seen as more “doable” than other types of qualitative designs.  With this type of design the researcher is able to use one site or one person, use a time limit, and use a limit on sources of data.  With this design the researcher follows a subject and uses observation, interviewing, videotaping, and review of related artifacts to collect their information.

2. Unstructured interviewing as a research design – this type of design is used mainly as a sole data collection technique by those researchers who have limited time and resources.  This design is free-flowing and unstructured/ semi structured.  Researchers keep clearly defined topics and sometimes questions in their mind to allow them to achieve the goals of their study.  The person being interviewed is told and topic and goal or the interview and then to direct the conversation.  Questions have to be asked in the same order in every interview, but they can be re-worded to fit each interviewee the best.  This type of research takes a great deal of practice Bailey says.  The interview should be tape-recorded to use as data later in the study. 

I would use a qualitative design if I was trying to learn about a single person or small group of people.  I would also use it if I want to study a group of people from a specific section of the world.  I would also use it if I wanted to learn about how something, such as a certain field in the medical world, developed over time.

10. Your instructor has stated that “…the best positivistic (quantitative) studies often arise from a relativist study or (qualitative) framework of inquiry.” Discuss advantages and disadvantages of qualitative and quantitative research designs.  If you had it to do over, would you change the methodology you used in your study?

Advantages of quantitative research are that the researcher can show causal explanations and relationships between variables, allowing for predictions.  The studies are predetermined and structured and do not change over time.  Therefore the researcher is able to use a defined operational model. The data are quantifiable and statistical.  The variables are defined ahead of time and are managed according to procedures ahead of time.  This type of data is much easier to manage than that of qualitative research. The sample size is very large allowing more randomization to occur to pick the subjects used. 

Disadvantages of qualitative research are that the researcher is unable to actually learn personally about the subjects that are being used in the study.  The predetermined structure of quantitative research can also be a disadvantage because if the study falls away from this predetermined structure or is found to not match it the researcher would have to start over or change form their original study. The researcher is very distant from the subjects and mainly observes.  I feel that this can lead to a misinterpretation of data at some point.  The tools used to collect data can be extremely complicated because the researcher may use scales, tests, and various types of hardware to complete the study. The studies here can be seen as not being relevant to real life and validity may be questioned. In this type of research the researcher may have trouble controlling variables too.

Advantages of qualitative research are that the researcher is able to sensitize readers to cultures and allow for an understanding of the perspectives of the subjects from a particular setting. The structure of this type of researcher allows for change to occur as the study progresses. The data here look at qualities.  I feel that this is a much more personal type of research than quantitative.  The tool for research is the researcher their self, which makes the data collection easier to do in most cases.

Disadvantages of qualitative research are that the researcher is unable to set a specific research question to be answered by the study.  The studies are way more generalized and can change at any point in time.  Because this type of research is flexible it may be hard for other researchers to follow a specific design.  I feel that it would be harder to complete this type of research.  The data is very hard to manage and requires specific technique which is harder to find since the data deals with qualities. Sample sizes are small which may be unrepresentative of the actual population.  The process can be very time consuming and yield large amounts of data which the researcher then has to place into length descriptive documents to conclude their study. 

If I was to do it over I would not change the methodology of my study.  I chose to do a non-experimental quantitative research design because I wanted to show the relationship between two variables.  I do not feel that a qualitative research design would have worked for the type of research I did and I also would not have been able to answer my specific hypothesis.  I needed a large sample to be representative of the population too, so quantitative research was my best choice.