OBESITY IN YOUNG PEOPLE

QUESTION

Background to the Lab Report

This lab report (a paper review) is quite different from those you may have done in the past.  It has been designed specifically for this course (designed by Reece, 2011) to complement and test your knowledge of research methods and data analysis. In the undergraduate psychology program you are exposed to a variety of different lab report formats. The aim is to help develop your skills and knowledge about the different ways that research is conducted and communicated in modern psychology—e.g., a qualitative study in social psychology, experimental study in cognitive psychology, a single-subject ABA intervention, at least one poster presentation, a survey evaluation, and so on.

In this lab, you are cast in the role of someone who is reviewing a journal article that has been submitted for publication. When manuscripts are submitted for publication, they are reviewed by at least two people using a system called blind peer reviewing. What this means is that you don’t know who is reviewing your paper, and the author’s name and affiliation is often removed from the manuscript so the reviewer does not know who has submitted the paper or what lab it comes from. This is an attempt to avoid all sorts of biases creeping into the review process.

What documents or files are attached?

Under Assignments on the Learning Hub you will find four files that form the basis of your Lab Report in Research Methods in Psychology.  The mock research paper (by John Reece) is the PDF file, designed to represent a “real” manuscript. It is obviously more basic than the research you would find in papers submitted for publication, but the basic structure and presentation is pretty much what you would see in a “real” manuscript. What is unusual is that you have been provided with the raw data (in SPSS and Excel formats) that formed the basis for the manuscript. This is rare, but it does happen. Also, there is a letter to the editor that will form the basis for your report.

What do you need to do?

Using the letter template, your job is to fill in the section after “My review is as follows”.  What you need to do is point out the flaws in design, method, analysis, and interpretation that has led to your strong rejection of the paper. That is, you need to (i) highlight what the author has done and where he has gone wrong, and (ii) correct his errors by providing the correct analysis of the data, including correct versions of the tables, if need be. Your focus must be on the methodological aspects of the manuscript. Don’t waste time and space criticising the format and wording of the introduction–that is not the purpose of the exercise.

There is no set format for your review, just as there is no single perfect format for the way “real” reviewers approach this task. You can use numbered or bullet points, or you can use sub-headings. In fact, part of the assessment will be based on the structure you apply to the task and how well that gets your message across.

The word limit for the report remains at 1000 words; that’s 1000 words of what YOU write, so it doesn’t include the text already provided for you in the mock letter to the editor.

This is a totally new report for you, and you may feel that you are swimming in the dark when it comes to presenting the report. This is where the Discussion Board comes into its own. Please use it freely to post any questions you might have, no matter how naive you think they might be. Make sure you use the discussion board as your main forum for posting questions.

SOLUTION

Introduction:

Many studies have been performed in order to identify the drivers for obesity in young people across the world. The author cites different studies that show a significant rise in the obesity levels of the youth population of different countries. The author through this paper claims to study the relationship between obesity and excessive watching of Television among young adolescents. The author attempts to test the following hypothesis:

1. Will high levels of television viewing directly cause an increase in weight?
2. Is there a different in the above tested relationship for male and female participants?

In order to test the hypothesis, the author performed a quantitative analysis of the survey responses from the participants.

Method:

• The study was performed with participants only from government schools which have different infrastructure and other facilities than a private school, thus, restricting the study only to a certain section of the population.
• The study was concentrated only on the northern region of the country the results of this study cannot be generalized for the population as a whole as the sample of the study lacks diversity.
• The sample size adds to the faults as it is too low to perform such a study. In the introduction the author has cited studies performed in this field and the sample size of those studies was much higher than this study.
• The whole method of forcing the participants to fill up the survey doesn’t add up as an effective methodology for performing such a kind of study.
• The author further limited the study to participants with their family names starting with letters from A-K. This particular method does not indicate any kind of significance to the research being performed.
• The Chi-Square Test performed to test the Goodness of Fit of the samples in order to decide if the bys are significantly older than girls could not be reviewed as the data provided did not provide any information about the age of the individual participants.

Sampling of the data is one of the important aspects in a research study which defines whether or not a study has been implemented using appropriate methodology. In the context of this study, the author has failed to perform a proper data sampling which implies that the results inferred by the author might be misleading.

Materials and Procedure:

Survey:

A research survey should contain self explanatory questions and have certain options from which the participant is either to choose one or more depending upon each individual’s choice. The author, however, has performed a survey in which the participants were given a Television guide. It would have been a well designed survey and also would have saved a lot of time if the author provided multiple choices for questions related to the recent shows they had watched in the survey. A well designed questionnaire forms the basis of the research as the type of questions (qualitative of quantitative) define what type of analysis needs to be performed. The very fact that the responses of all the participants have been scored by student themselves (randomly distributed) makes the process unconventional and disorganized as it is the researcher who should be dealing with the data rather than the participants even if they are secondary school students.

Results:

The author had performed exploratory analysis on the data using SPSS software in order to find any errors as well as to test the normality and homogeneity of variance. However, the author did the following mistakes:

• Histogram: It is used to plot the frequency distribution, check the normality of the data and also to categorize the given data into certain groups depending upon the distribution (Tulane University). But the author claims to inspect histogram in order to check the errors in the data which would not be possible through this method.
• Stem and leaf plot: It is another way of plotting the data in the form of a histogram (san jose state university) and, thus, cannot be used to detect errors in the data.
• Normality Plots, Kolmogorov-Smirnov and Shapiro-Wilks tests: The normality plot is a means of plotting the data in order to check whether or not the data is normally distributed (Ryan, Jr., TA and Joiner, BL, 1990). It is one of the alternatives to the Kolmogorov-Smirnov and Shapiro-Wilks tests which are performed to examine the normality of the data. Therefore, the author of this study has used the test and inferred the results from these tests incorrectly and for inappropriate purpose.
• Levene’s test:  It is a test used to check the homogeneity of variance of the data (Pennstate), unlike the author who tried to check the normality of the data with this test.

These faults in the research indicate that the author has been compromised with the results which are both wrong and incorrectly interpreted.

The author further claims that the variables exhibit high levels of positive and negative skew in the test of normality, which is wrong as the normality tests performed on the variable sex provide the following table:

It can be inferred from the above table that in both the cases the values are not significant enough to indicate normal distribution of the variable. Therefore, no information regarding skewness of the variable is indicated by the normality tests as claimed by the author.

The University of Homogeneity of Variance was tested using the independent samples test which is shown below:

The significance values of all the values are higher than 0.05, which means that the hypothesis of homogeneity of variance is rejected; unlike the author’s method (using the normality tests to cheek homogeneity of variance) we used the Levine’s test to reach at this conclusion.

The table below shows the descriptive statistics of all the variables:

The above table shows that the variables exhibit low levels of positive and negative skew but the variables Sex and TV Viewing Hours have high negative values of kurtosis which are undesirable. This shows that the author’s method as well as the results obtained are wrong with respect to the kurtosis and skewness.

The table above shows the descriptive statistics for the male participants and the table below shows that of the female participants:

The author mentions that these descriptive statistics indicate statistically significant differences between the responses from male and female respondents, however, the tables shown above were obtained from the SPSS and no significance values were displayed, therefore, no such inferences can be made through these statistics.

The tables below show the correlations between different variables for the total sample, male participants and female participants respectively:

Total Sample:

Male Participants:

Female Participants:

All the three tables show high correlation between the following:

• Weight & TV Viewing Hours
• BMI & weight
• TV Viewing Hours & BMI

However, the correlations are found to be higher in case of female participants over male participants.

The paired sample test is generally performed on normally distributed variables in order to check to see any differences between two related observations (UCLA). In this case the author has tried to check the sex differences for different variables using paired sample test which, we believe, is not the correct test to perform.

The results from the Descriptive statistics only show the mean and standard deviation of the different variables but do not talk about the significant differences between the male and female participants, as the author has stated wrongfully.

Discussion:

The author’s hypothesis of strong and significant causal relationship between the BMI, weight and television viewing hours cannot be concluded as the data is neither normally distributed nor there is homogeneity of variance in the data, however, the higher correlation coefficients between these three variables do indicate some kind of relationship among them. The author also mentions that the previous studies performed have indicated a strong relationship between passive forms of entertainment and overweight among young adolescents. The author based on the results from the research recommends limiting the television viewing in order to deal with the obesity issues found in the young adolescents. As already stated the recommendations are clouded by wrong interpretations and results from the research methodology followed by the author.

The conclusion regarding the sex differences in the relationship between different variables are also inappropriate and would wrongfully lead the readers to take actions which might not be necessary. The author’s biasness towards the effect of TV viewing hours on the obesity of female participants is not backed by significant statistics and, thus, should not be considered.

To conclude, this research paper has incorporated many wrong methods to perform the study and the author makes inappropriate inferences and recommendation based on the faulty results. Therefore, the research paper should not be accepted.

References

Ryan, Jr., TA and Joiner, BL, 1990, Normal Probability Plots and Tests for Normality. [online] Available at: <http://www.minitab.com/uploadedFiles/Shared_Resources/Documents/Articles/normal_probability_plots.pdf> Accessed on 7 May 2012

Tulane University, Chapter-3 One-way Analysis. [online] Available at: <http://www.tulane.edu/~panda2/Analysis2/one-way/pg1%20descriptive%20analysis/oneway01.htm> Accessed on 7 May 2012

San Jose State University, Frequency Distributions. [online] Available at: <http://www.sjsu.edu/faculty/gerstman/StatPrimer/Freq.PDF> Accessed on 7 May 2012

Pennstate, Levene’s Test of Homogeneity of Variance. [online] Available at: <http://www.personal.psu.edu/users/d/m/dmr/papers/levenes.PDF> Accessed on 7 May 2012

UCLA, Statistical analyses using SPSS. [online] Available at: <http://www.ats.ucla.edu/stat/spss/whatstat/whatstat.htm> Accessed on 7 May 2012

KF20

Please  Click on the  below links to Chat Now  or fill the Order Form !