nging the "profile" of the paper. They are now interested in finding out if the distribution over educational levels among readers has changed since the change in the profile. You have been given the task of analysing if the distribution has changed.
adolescents. I have completed the Literature review. I have to plan a research design for the thesis. ABOUT THE THESIS:
Population- 13+ to 19 years adolescents
No of IV s- 2 that is, Gender is the First IV and Locus of Control ( LOC) is the second IV
Levels of IV- Gender has 2 levels ( male and female). LOC has two levels ( Internal Locus of Control and External Locus of Control). Both IV s are categorical variables.
List of DV s-
1. Self Esteem
4. Academic motivation
5. Exam Anxiety
6. Life Style
Each DV will be measured by using appropriate Statistical scales. All DVs will be taken as continuous variables. All the scales measure the quantitative aspect only.
DV MEASUREMENT: Each of the scales that will be used to measure DV contains several areas or dimensions or sub-categories. For example, the self-esteem scale contains 78 items divided into 6 categories like personal, social, emotional, academic, intellectual and moral.
THESIS AIM- To check the impact of IV s on each of these DV s in isolation and also investigate the interaction effects between gender and locus of external. I will be using the SPSS package for calculations.
Problem 1: Which is the most appropriate Statistical test or design that should be used here? I believe a 2x2 ANOVA will be best suited here.
Problem 2: If I am measuring the impact of IV s on each DV in isolation, should I use several Two Way AONVA tables or a single MANCOVA table?
Problem 3: Each of the DV is measured using scales containing several dimensions. Are such dimensions of the scales need to be treated as the levels of the dependent variables? In other words, do the levels of DV are decided as per the dimensions of the scale that was used to measure the DV? If this is so, then even if I am measuring the DV s in isolation; each DV will have multiple levels, which in turn will change my design from Two Way ANOVA to perhaps MANCOVA? What is the right approach here?
Note: I have not intentionally divided DV into any levels.
Kindly help me to arrive at a statistically significant research design! If possible, kindly briefly explain the type of the design as well as the rational or suitability of the sign for my research problem.
est, correlated groups t test, 1 way bet-sub Anova, 2-way bet-sub Anova, 1-way rep-measures ANOVA or test for signif. of correlation.
Attachment theory hypothesizes that children's relationships with their parents shape habitual patterns of attachment in close relations. Thus, ones childhood experiences should connect to relationship quality during adulthood. Many years ago, a large group of teenage boys had been interviewed about their families. These interviews were coded to provide data in childhood family environment- a measure on which low numbers reflected distant/punitive/cold relationships with parents while higher numbers reflected nurturing/autonomous/warm relationships. Many decades later, researchers attempted to re-contact these same people. A total of 81 men(aged 70-85 years) who had been married to the same spouse for at least ten years and agreed to participate were interviewed about their marriage. The spouses of these men were interviewed as well. The interviews were coded in order to provide a measure of relationship attachment-higher scores reflected greater levels of relationship satisfaction,loving, and caring. Is there any evidence to support attachment theory's claim that children's parental relationships predict relationship quality in adulthood?