# Exercise 1: Data Management

Irian Joseph Biag, Sean Hirshel Cusi, Juwan Eugenio, John Angelo Tan
Group 6
Introduction
It is important to have a firm understanding of data analysis as a researcher; it requires a
great deal of work and care to write an accurate and informative research. A researcher must then
effectively convey the data he has obtained and the ideas he has come to. Before the process of
dissertation writing, researchers must first collect data for the study. The data to be used can
come by using data gathering techniques, like sampling for instance, or by using data from other
researches if it will help you prove a point. There are long and arduous steps to collect and
analyse data but these are necessary to procure accurate and correct data for any research. Data
analysis is vital and meticulous as it gives an explanation of various concepts, theories,
framework and methods used in the research and it helps the researcher arrive to the conclusion
and thus proving and disproving the hypothesis.
Data analysis is the process of assessing data using analytical and logical reasoning to
examine in each component of the data provided. One type of analysis where you can evaluate
between the difference of before and after observations on the same subject or compare two
different methods of measurement or where the measurements are applied to the same subjects is
possible. Paired T-test is applied preferably. A paired t-test is used to compare two population
means where you have two samples in which observations in one sample can be paired with
observations in the other sample. It is a parametric procedure, where it makes several
assumptions. These are; Dependent variable must be continuous, normally distributed, does not
contain any outlier, and observations are independent with each other. Lastly, There are two types
of significance to consider when interpreting the results of a paired sample t-test, statistical
significance and practical significance. Statistical significance is determined by looking at the p-
value. Practical significance depends only on the subject matter.
To analyze different sets of data, researchers must know how to use different statistical
methods. An example of this is, Wilcoxon signed-rank test. Wilcoxon signed-rank test is an
alternative to the paired t-test, used when the data gathered does not meet the requirement
(normally distributed) for the paired t-test. It uses ranked or ordinal data. The same with paired t-
test, Wilcoxon signed-rank test is used when there are two nominal variables and one
measurement variable. Aside from analyzing the signs of the differences, the magnitude of
observed differences is also considered in this test.

Results:
Problem: A species of non-native fish was accidentally introduced into a lake where an
endangered native fish species is found. Although the introduced species does not attack or eat
the native species, it does have the same dietary preference as the native fish. Scientists, afraid
that the introduced species may be indirectly out-competing the native species for food, decided

0 12.1 3 10.to do a short-term experiment.7 13.3 12.  Biological Ha – The non-native fish out-competed the native fish for food.2 2 11.9 7 12.4 8 12. the scientists started trapping all the fish in the lake until they located the exact same fishes that they weighed three months earlier.3 12. Fish# Original +3 months Weight(kg) Weight (kg) 1 10.9 12. Biological Hypothesis:  Biological Ho – The non-native fish did not out-compete the native fish for food. weighed each one of them and tagged each fish with its own unique ID number (1-10). They collected 10 samples if the native fish species. After 3 months. They took the new weight of the fishes and compared these with their new weights three months before.5 Question: The scientists want to know whether the introduced non-native fish indirectly out- compete the native fish for food.9 5 10.0 6 11.2 11.9 12.4 12.0 11.5 10 12.1 9 11.0 12.1 4 12. Independent variable: the native fish Dependent variable: the weight of the fish Statistical Hypothesis:  Ho – There is no significant difference between the medians of the paired observations  Ha – There is a significant difference between the medians of the paired observations Variables: 2 nominal – weight and the individual native fishes 1 ordinal – weight values in kg Test: Wilcoxon Signed Rank Test .

the non-native fish does not indirectly out-compete the native fish for food. The test ranks the differences according to size and is therefore a non-parametric test. the dependent samples must be of ordinal nature.049). meaning there is always randomness and independence when drawing. or should we say a “before and after” measurement.The Wilcoxon sign test makes four important assumptions. in this case the weight of the native fish and the weight of native fishes 3 months later.. Therefore. (2) Independence. wherein the original weight is not equal or greater than the after 3 month weight. there is a statistically significant difference (P = 0. Conclusion: There is a significant difference between the original weight and the weight recorded after 3 months. Inference: The change that occurred with the weight recorded after 3 months is greater than the expected by chance. (3) Continuous dependent variable. (4) Ordinal measurement. For Problem F: Statistical Hypothesis:  Ho – There is no significant difference between the mean of the paired observations  Ha – There is a significant difference between the mean of the paired observations . But in this research the samples were deliberately searched for because the research is finding the original fish to record their growth. (1) The dependent samples. so that we can compare them.

Therefore. the native fish actually became heavier and proves that the non-native fish does not out-compete the native fish for food. Rejecting the hypothesis that the sample mean of the original weight is equal to or greater than the sample mean of the weight recorded after 3 months. o Ha1 – the sample mean of the original weight is greater than the sample mean of the weight recorded after 3 months.the sample mean of the original weight is lesser than the sample mean of the weight recorded after 3 months. Test: Paired t-test Inference: The sample mean of the weight recorded after 3 months exceeded the sample mean of the original weight. o Ha2 . Conclusion: Results show that the sample mean of the weight recorded after 3 months is higher than the sample mean of the original weight. .