|← Bowling for Columbine Propaganda||Response to Questions →|
A hypothesis is generally a prediction one makes about what is likely to happen. In many research situations, a hypothesis is an informed educated guess based on logic, theory, or past data. A hypothesis must be consistent with past researches, concise, testable and predict the relationship between variables.
When designing an experiment, the research hypothesis may be directional or non-directional. For this discussion, identify the differences between a directional and non-directional hypothesis. In addition, discuss how the directionality is related to the one- and two-tailed t-test. Is there an assumption that is violated if a two-tailed t-test is used with a directional hypothesis? Explain your reasoning.
Research starts by stating the null hypothesis. The null hypothesis is a beginning point of testing whether the hypothesis is true. There is also an alternative hypothesis that states that the hypothesis is wrong. Therefore, research hypotheses are alternative and null hypothesis (Toledo, Flikkema & Toledo-Pereyra, 2011). The hypothesis tested can be either directional or non-directional. A non-directional hypothesis is a two-tailed test where alternate values lie on both sides of the null hypothesis. A non-directional hypothesis predicts a change of the dependent variable if the independent variable is manipulated, but not the direction of the change. It states that a correlation exists but it is not clear whether the relationship is positive or negative. The directional hypothesis is a one-tailed test. The alternate values lie on one side of the value specified in the null hypothesis. A directional hypothesis states that either a positive or negative correlation exists between variables. Therefore, it predicts the direction in which dependent variable will change if the independent variable is manipulated.
A t-test is any hypothesis used to test statistics. The directionality is related to one and two tailed t-tests. In a two-tailed t-test, if one is using a 0.05 as the significance level, half of one’s alpha goes in one direction, while the other half goes in the opposite direction. This shows that the probability of the relationship lies in both directions. In one-tailed t-test, if one is using the same significance level of 0.05, one’s entire alpha goes in one direction of statistical interest. This shows that the probability lies in one direction.
According to ExoAnalytics Inc (2011), the assumption that is violated states that the variances to be used in t-tests must be equal. This is because for a one tailed t-test, the alternative hypothesis is that the difference between standard deviation is less or greater than the proposed value. In the two tailed t-test, the difference between standard deviations of the alternative hypothesis is not equal to the proposed value. Improper use of statistical tests leads to invalid results.
For this discussion, briefly design an experiment that uses a repeated measures design. The brief design should explain the research question, the subjects, and the variable(s) of interest. What method would you use to analyze the data from this experiment and why? Indicate whether or not the analysis would contain any type of post hoc comparison. Be sure to mention the subject of power and the analysis of power in your discussion of your subjects.
Impact of Caffeine on Cognitive Functioning
Age, consumption levels of caffeine, and one’s lifestyle have a placebo effect to prior consumption of caffeine. The study tests the existence of confounding variables like lack of sleep using two aspects that influence caffeine consumption. The research question of interest is whether caffeine leads to high heart-rates while slowing down cognition among participants, who mistakenly believed they had consumed caffeine. The experiment aims to examine dependent variables separately in two groups (people who consumed caffeinated and decaffeinated drinks) to examine whether the placebo effect existed in the sample.
Subjects of this study are 16 college students divided into two groups of 8 people. One group drinks caffeinated coffee while the other drinks decaffeinated coffee randomly during the day. Subjects are not aware about which participants take caffeinated or decaffeinated coffee. The subject power is the sample size that is used to detect the effect. Subject variables collected are those that are believed to have an effect on cognition after consumption. The randomization of subjects for pre-test and post-test design is a control for personal differences, which stands as the analysis power.
The experiment uses online questionnaires because it gives a possibility to present participants with a description of the project and informed consent forms. Participants are asked to calculate their heart-rate and record it in the questionnaire. They have to take 6 cups of provided coffee a week. They must also test their heart-rates 10 seconds after having a drink daily and record it as well.
Paired t-tests between the post-test and pre-test conditions vary in each group. This experiment’s analysis power assumes that randomization reduces threats of validity. Statistical analysis shows variance in participants’ heart rate before and after the consumption of drinks. This is how the analysis contains post hoc comparisons. The manipulated independent variable is caffeine drink, which shows that the group which consumed caffeine experienced an increase of heart rate to a significance level p value of 0.6. The control group also showed some change in heart rate due to placebo effects (p value of 0.4).
Heart rate and cognition change significantly after the consumption of caffeinated coffee. The analysis shows that heart rate and cognition change also after consumption of decaffeinated coffee due to person’s perception i.e. placebo effect.