Psychology Research Designs - Annotated Bibliography Example

Running Head: PSYCHOLOGY RESEARCH PSYCHOLOGY RESEARCH DESIGNS Psychology Research Designs Basic Concepts/Terms in Statistics Used by the Journal Articles Being Reviewed In all the articles that were studied, there are several terms in statistics that kept coming up. Thus, it was deemed important to understand what these common terms mean. 1. Hypothesis–simply a conjecture about the value of a certain parameter or the distribution of values for a parameter a. Null hypothesis (H0 or H0) – the statement or conjecture that is being tested in the hypothesis. It is usually a statement that a certain treatment has no effect. Most experiments make use of a null hypothesis although the common objective is to prove the null hypothesis to be false. b. Alternative hypothesis (HA or Ha) – the complementary statement that will be accepted once the null hypothesis is proven false and rejected. 2. N or n refers to the number of individuals or samples from which measurements were taken. 3. Mean–the value which represents the totality of measures divided by the number of samples from which the measures were taken. In simple terms, the average value 4. Range–a quick and rough measure of variability. It is the difference between the minimum and maximum values of a data set. 5. Standard deviation (sd or SD or s.d.)–this is the most common measure of variation and it measures how different measurements or values are spread around the mean. If it is large, it means that the data is widely spread about the mean and if small, then the data are close to the mean. 6. Standard Error of the mean–measures the spread of the error in the process from where the mean is obtained 7. Variance–it is derived from the standard deviation and calculated by multiplying the sd by itself. 8. P value (p)–measures how much evidence there is against the null hypothesis. Researchers will usually reject a hypothesis if the p-value (also denoted as α) is less than 0.05 or 5%. This means that the probability of the null hypothesis being true is 5% if one is to repeat the experiment. 9. F value–this is also called the variance ratio. As p value decreases, the F-value increases, a higher F value will show that there is difference between or among treatment groups. 10. Population–set of all possible values of a certain variable; for example population for sex would be female and male Use of Statistics in the Research Designs in Journal Articles Selective Serotonin Re-Uptake Inhibitor Treatment-Emergent Sexual Dysfunction: Randomized Double-Blind Placebo-Controlled Parallel-Group Fixed-Dose Study of A Potential Adjuvant Compound, VML-670 (Baldwin et al., 2008). The paper repors on the use of SSRI, a class of drugs, to treat sexual dsysfunction. Patients or subjects of the experiment were randomized or randomly assigned to groups in a double-blind treatment with placebo (contrrol group) or VML-670. To assess the response to VML-670, the ASEX scale was used. Statistical analysis made use of the ITT-LOCF methods to take care of missing data when some of the patient ended treatment or left the experiments. 1. Hypothesis to be tested–that VML-670 would prove efficacious in treating sexual dysfunction due to paroxetine or fluoxetine. 2. Double-blind treatment–in a double-blind experiment neither the researchers nor patients/subjects know which treatment group (placebo or VML-670) the subjects belong to. This is a stringent way of eliminating bias. 3. Arizona Sexual Experience Scale (ASEX)–this refers a 5-item rating scale that measures in quantifiable terms elements of a sexual experience which could indicate sexual dysfunction. Thus, a higher score (maximum 30) means more sexual dysfunction. 4. Intention-to-treat, last-observation carried forward analysis (ITT-LOCF)–the intention to treat analysis includes all individuals or patients randomly assigned to groups regardless of adherence to entry criteria, treatment received and whether they subsequently deviated from the protocol or left the trials. Withdrawals and deviations from experiments are included because these happened after randomization or grouping and such activity can be due to the treatment itself. When this happened in Baldwin et al. (2008), the researchers made use of data before the withdrawals occurred (last data without the missing data). 5. Wilcoxon two-sample test–used to measure the difference in two treatments. Used for variables that are non-parametric (or non-numerical) 6. Chi-square test–similar to the Wilcoxon two-sample test, the chi-square test will tell you if the difference between the two treatments are significantly different. Comorbid Depression among Untreated Illicit Opiate Users: Results from a Multi-site Canadian Study (Wild, et al., 2005). In this study, patterns of major depression (MDD) were observed in a group of illicit opiate users from 5 locationsin Canada. A survey was conducted on 679 users; the data included health service use, drug use, sociodemographics and depression. From these, logistic regression analysis was conducted to relate to the occurrence of MDD. Chi-square analysis was used to compare the injection drug and cocaine useamong participants with and without depression. 1. Logistics Regression Analysis–it is an analysis used to predict a response based on several measured variables. In Wild et al. (2005), logistic regression was used to predict the occurrence of depression. The dependent variable here is occurrence of MDD and the predictors are sex, ethnicity, housing site and recruitment status. Logistic regression was also used to identify independent predictors of depression: being white, being a woman and living outside Vancouver. 2. Odds Ratio–measure of the chances of an event occurring in one treatment group over the odds of that event occurring in the other treatment group. 3. KruskalWllis Test–a measure of differences among or between treatment groups based on nonparametric or non-numerical variables. 4. Chi-square Test–also used to measure differences among nonparametric variables. Serotonin Transporter Occupancy Induced By Paroxetine In Patients With Major Depression Disorder: A 123I-ADAM SPECT Study (Catafau, et al., 2006). The paper discusses the effect of paroxetine, a selective serotonin reuptake inhibitor antidepressant on the serotonin transporter occupancy (SERTocc) in patients with MDD using 123I-ADAM SPECT, an equipment which can be used to visualize SERTocc. ADAM SPECT readings were taken from MDD patients before they started paroxetine and 4-6 weeks after receiving the drug. The readings were compared to those from healthy volunteers (control group). A one-way ANOVA was used to compare the SERTocc in different regions of the brain and also to compare pre-treatment specific uptake ratios. Paired t-tests were used to compare the specific uptake ratios in pre-treatment patients and helathy volunteers. 1. One-way ANOVA–ANOVA is analysis of variance; it is a test used to compare variations in response to treatments or in this paper, the different regions in the brain.. One-way ANOVA is the simplest of the tests because it compares only one factor (e.g. effect of paroxetine on SERTocc). 2. Paired t-tests–it is used to compare data between groups. In the paper, it was used to measure specific uptake ratio in healthy volunteers and drug-free patients 3. SPSS or Statistical Package for the Social Sciences–it is a software used to analyze data that are gathered in studies of the social sciences, medical field, education, to name a few. Catafau, et al., (2006) used SPSS to come out with their ANOVA and t-tests. SSRI Use and Bone Loss in Older Women (Miller, 2008.) The study reports on the the relationship between the use of SSRI and bone loss in older women. It summarizes the evaluaton of osteoporotic fractures in women from 4 areas. The evaluation lasted for more than three years. However, the paper did not present much on the statistics used to anallyze their data but it does show the level of significance or P value when they compared bone mass density between SSRI users and non-users. 1. P value–probability value which when it is above 0.05 or 5% shows that the treatment did not result in significant changes. Lower P value will show more effect of treatment. Stressful Life Events, Genetic Liability and Onset of an Episode of Major Depression in Women (Kendler, et al., 1995). The study is on the effect of genetic liability (or genotype) and stressful life events lead to the onset of depression. A population based sample of female twins were invited for interviews twice (17.8 months apart). The research was able to categorize the different ‘stressful’ life events which lead to depression and the effects of genetic makeup on the onset of depression. Linear and logistic regression analyeis were used to correlate stresses and genotype to the onset of depression. 1. Population based sample–this means that from a large population, a small subset or number of people were chosen to participate in the study or survey 2. Linear Regression Analysis–used to predict a response based on measured variables. Results of this analysis are quantitative and shows the direct relationship between two variables. 3. Logistic Regression Analysis–it is similar to linear regression analysis but it makes use of several variables to predict a categorical response where the relationship is linear. This was used to determine the relationships between stressful life events and environment, stress and genetics, stress and genetics and environment, all in bringing on the onset of depression. 4. Odds ratio–ratio of the odds of an event occurring in one group over the odds of that same event occurring in the other treatment group. 5. Algorithm–a computer program used to solve for any given problem. 6. χ2 (chi-square) (also called Pearson chi-square test) – shows result of the chi-square test or chi-square distribution which tells if variation in the data is present. If the results are significant (denoted by * or **) then the data deviates or the effect of treatment varies from the expected. 7. df (degrees of freedom)–this shows the number of variables that can vary. For example if there are two variables, then only one variable can vary or the df =1. Genetic Markers of Suicidal Ideation Emerging During Citalopram Treatment of Major Depression (Laje, et al., 2007). The study reports in the identification of genetic markers that are related to suicidal ideation during treatment for depression. In the study, outpatients with depression were treated with Citalopram. DNA was collected from them and genotype for single nucleotide polymorphisms (or simply, differences in DNA sequences) for 68 candidate genes. One-hundred twenty patients developed suicidal ideation in the course of the treatment and their allelic frequencies were compared to the other patients (control) in the study. Two genetic markers (or gene sequences) were identified to be associated with suicidal ideation. Statistical tests centered mostly on mean comparison and presence or absence in variation. 1. Kolmogorov-Smirnov Test–This test simply compares the means or values in one group with another group that has a normal distribution (treated versus control group). 2. Kruskall-Wallis Test–measures non-parametric differences between two groups 3. Covariate –Factors that affect the individuals or samples as a whole. For example, height, weight, body mass index (BMI). 4. χ2 (chi-square) (also called Pearson chi-square test)–shows result of the chi-square test or chi-square distribution which tells if variation in the data is present. If the results are significant (denoted by * or **) then the data deviates or the effect of treatment varies from the expected. 5. Hardy-Weinberg equilibrium–describes the state where there are no changes in the population and the genetic make-up or constitution remains the same throughout evolution. 6. Hosmer-Lemeshow Test–this test shows the goodness of fit of a derived model (based on the results of regression analysis). If results are non-significant, then the derived model has a good fit. The Epidemiology of Major Depressive Disorder: Results From The National Comorbidity Survey Replication (NCS-R) (Kessler, et al., 2003). The paper seeks to study the prevalence of major depressive disorders (MDD) in the United States of America using the National Comorbidity Survey Replication which is a face-to-face household survey. The data gathered included estimates of prevalence, age-at-onset of MDD, treatment, sociodemographic distribution of comorbidity. The following are the tools used to assess the epidemiology of MDD in the United States. 1. Confidence interval (CI)–interval estimate of a measured parameter. This is used to estimate the reliability of estimations. A smaller CI will show that the estimate is more reliable than a higher CI. 2. Cross-tabulation–it is a method of showing the joint distribution of 2 or more variables in a matrix or table form. 3. Kaplan-Meier Test–it is used to estimate survival or life or time function. In Kessler et al. (2003) the Kaplan-Meier was used to estimate age-at-onset of major depressive disorder (MDD). 4. Logistic Regression Analysis–it is a model used to predict a response based on several measured variables. In Kessler et al. (2005), logistic regression was used to study the correlation of prevalence of MDD and treatment. 5. Multivariate Significance Tests–similar to other regression analysis but there are two or dependent variables. In the paper, the Wald χ2 Test or a chi-square test was used to determine the significant effects of the different covariates on the prevalence of MDD. 6. Taylor Series Linearization Method–a method for linearization of values in a non-linear relationship. This is used to understand linear or direct system in a non-linear function. References Baldwin, D., Hutchison, J., Donaldson, K., Shaw, B., & Smithers, A. (2008). Selective serotonin re-uptake inhibitor treatment-emergent sexual dysfunction: randomized double-blind placebo-controlled parallel-group fixed-dose study of a potential adjuvant compound, VML-670. Journal of Psychopharmacology, 22 (1), 55-63. Catafau, A. M., Perez, V., Plaza, P., Pascual, J.-C., Bullich, S., Suarez, M., et al. (2006). Serotonin transporter occupancy induced by paroxetine in patients with major depression disorder: a 123 I-ADAM SPECT study. Psycopharmacology, 189, 145-153. Kendler, K. S., Kessler, R. C., Wlaters, E. E., MacLean, C., Neal, M. C., Heath, A. C., et al. (1995). Stressful Life Events, Genetic Liability and Onset of an Episode of Major Depression in Women. The American Journal of Psychiatry, 152 (6), 833-842. Kessler, R. C., Berglund, P., Demler, O., Jin, R., Koretz, D., Merikangas, K. R., et al. (2003). The Epidemiology of Major Depressive Disorder: Results From The National Comorbidity Survey Replication (NCS-R). Journal of the American Medical Association, 289 (23), 3095-3105. Laje, G., Paddock, S., Manji, H., Rush, A. J., Wilson, A. F., Charney, D., et al. (2007). Genetic Markers of Suicidal Ideation Emerging During Citalopram Treatment of Major Depression. American Journal of Psychiatry, 164 (10), 153-1538. Miller, K. E. (2008, January 15). www.aafp.org/afp. Retrieved September 18, 2008, from American Family Physician: http://www.aafp.org/afp Wild, T. C., El-Guebaly, N., Fisher, B., Brissette, S., Brochu, S., Bruneau, J., et al. (2005). Comorbid Depression Among Untreated Illicit Opiate Users: Results from a Multi-site Canadian Study. Canadian Journal of Psychiatry, 50 (9), 512-518. Read More