Rasmussen College Essentials of Social Work Quantitative Data Analysis Discussion Response 1 1) Provide an example of each of the follow levels of measure

Rasmussen College Essentials of Social Work Quantitative Data Analysis Discussion Response 1

1) Provide an example of each of the follow levels of measurement: Nominal, Ordinal, and Ratio.

Nominal measurements are unrelated types of measures that do not relate to one another (Radey, 2020). For instance, in a study of dogs, nominal measurements could include the number of other pets in their home, color of fur, age, and breed. These are all characteristics that do not have any dependence on one another. Ordinal measurements can be ranked or put into order, but the difference in their positions is unknown (Radey, 2020). For example, a scale of how a person is feeling (Depressed to Extremely Happy) or the line of people waiting to use the bathroom. Lastly, ratio measurements can also be ranked or put into order, but unlike ordinal attributes, ratio attributes are evenly spaced (Radey, 2020). Ratio measurements are counted starting at zero (Radey, 2020) For instance, how many people standing in line during the COVID-19 pandemic who are standing on marks on the floor an even six feet apart or the number of COVID cases at a hospital in a single day.

2) Explain the difference between a univariate analysis and a bivariate analysis.

In univariate analyses, researchers examine cases one variable at a time (Rubin & Babbie, 2017). For instance, if natural color of hair was being measured, researchers would look at how many participants have brown hair, black hair, blonde hair, and red hair. These are all within one singular variable- the natural color of hair of the sample. In bivariate analyses, rather than measuring one singular variable, researchers study the relationships among multiple variables (Rubin & Babbie, 2017). Rather than looking at two variables separately, a bivariate analysis focuses on their relationship for explanatory purposes whereas univariate is for descriptive purposes (Rubin & Babbie, 2017). For example, explaining one group via a bivariate analysis is saying something like half of the sample is female and all of them have one child, while the other half are all male and have zero children. This shows a strong relationship between the variables of gender and number of children. This bivariate analysis can provide a plausible explanation for why women utilize a child service agency more than men.

3) You have provided a 15-session psycho-educational support group intervention for people who recently completed inpatient treatment for substance abuse. Of the 10 clients who were in the group, two attended all 15 sessions, two attended 14 sessions, two attended 13 sessions, two attended 12 sessions, and two attended just 1 session. Calculate the mean and the median number of sessions attended. Which measure of central tendency, the mean or the median, do you think is more appropriate to use in portraying the average number of sessions attended for the above data? Why?

Mean: 11

Median: 13

The more appropriate central tendency in this data is the mean, because there is a significant outlier in the data. Eight of the ten participants attended 12-15 sessions, but two attended just one session. The mean takes the outliers into consideration while the median only looks at the overall “middle” number when all of the data is written out.

4) Explain what the null hypothesis is and please explain how statistical significance refutes chance.

The null hypothesis, or rival hypothesis, says that the relationship being studied exists by chance (Rubin & Babbie, 2017). The null hypothesis explains a relationship as occurring by chance rather than actually existing within the population in question. However, when a study’s findings are shown to be statistically significant, the null hypothesis is proven incorrect (Rubin & Babbie, 2017). Statistical significance looks at the results of a study to determine how likely it is that the null hypothesis is true. It is important for researchers to prove a null hypothesis wrong to strengthen the plausibility of their findings.

5) Explain the difference between a Type 1 Error and a Type II Error.

When making determinations about the statistical significance of relationships, researchers can make errors. There is always a chance that null hypotheses are plausible, so they can never be called impossible, just significantly caused by chance (Rubin & Babbie, 2017). A Type 1 Error occurs when researchers reject the null hypothesis as implausible but it is actually true (Rubin & Babbie, 2017). There is no way to know this type of error is being committed, because there is always a chance, even if extremely slim, that the null hypothesis is true. While a Type 1 Error falsely rejects a null hypothesis, a Type 2 Error does not reject a false null hypothesis (Rubin & Babbie, 2017). In Type 2 Errors, research results are not significant enough to prove that the null hypothesis is false so chance plausibility is still possible.

6) Explain why statistical significance does not tell us the magnitude of the relationship between the independent variable and the dependent variable.

Statistical significance means that there is a good chance that we are right in finding that a relationship exists between two variables. However, we can never be completely certain that a relationship exists between two variables. There are too many sources of error to be controlled. Therefore, it is difficult to determine the magnitude of a relationship that we are not 100% certain actually exists. To say an independent variable causes a change to a dependent variable at a certain magnitude, we have to be certain no other variables are impacting the change. There are too many sources to be controlled to determine the exact level of change caused solely by the independent variable. For example, social support (IV) decreases depression (DV), but we do not know the magnitude of this relationship because everyone experiences different results, which could be due to outside variables like other treatments and medications, relationships with family and friends, career, age, gender, life experiences, mental health history, and such.

Megan Evans

References

Radey, M. (2020). Analyzing data. [Powerpoint slides]. https://canvas.fsu.edu/courses/128586/files/823796…

Rubin, A. & Babbie, E. (2017). Research methods for social work (9th ed). Independence, KY: Cengage.

Response Pose 2

Provide an example of each of the following levels of measurement: Nominal, Ordinal, and Ratio.

Nominal: Which is your preferred dessert? (Categories with no ranking or direction)

Key Lime Pie
Chocolate Cake
Pudding
Ice Cream
Other

Ordinal: I am satisfied with the services I received today. (Categories with ranking)

Strongly Agree
Agree
Neutral
Disagree
Strongly Disagree

Ratio: What is your age? (Difference between measurements)

Under 21
22-29
30-39
40-49
50+
Explain the difference between a univariate analysis and a bivariate analysis.

A univariate analysis is the analysis of one (uni) variable to describe. Bivariate analysis is the analysis of two (bi) variables to determine and explain the relationship between them (Rubin & Babbie, 2017).

You have provided a 15-session psycho-educational support group intervention for people who recently completed inpatient treatment for substance abuse. Of the 10 clients who were in the group, two attended all 15 sessions, two attended 14 sessions, two attended 13 sessions, two attended 12 sessions, and two attended just 1 session. Calculate the mean and the median number of sessions attended. Which measure of central tendency, the mean or the median, do you think is more appropriate to use in portraying the average number of sessions attended for the above data? Why?

Median: 13 Mean: 11

In this hypothetical situation, I believe the median (middle) number of sessions attended by the clients would be the best portrayal of data because the data is skewed with some clients attending many of the sessions and a few clients only attending one. Rubin & Babbie state it is best to use a median number when the data is presented as in this scenario.

Explain what the null hypothesis is and please explain how statistical significance refutes chance.

A null hypothesis is a hypothesis that suggests there is no relationship between the variables that are being tested. The researcher may determine that the two variables are related after rejecting the null hypothesis (Rubin & Babbie, 2017).

According to Rubin & Babbie, statistical significance refers to the unlikelihood that the relationships observed in a sample are due to sampling error alone. Statistical significance refutes the chance if the p-value is less than 5%. If the p-value is greater than 5%, then the results are due to only chance and the data are deemed to be consistent with the null hypothesis. If the p-value is less than 5%, the data are considered to be statistically significant and inconsistent with the null hypothesis (Rubin & Babbie, 2017).

Explain the difference between a Type 1 Error and a Type II Error.

A type 1 error is an error we risk committing when we reject the null hypothesis. A type II error is an error we risk when we fail to reject the false null hypothesis (Rubin & Babbie, 2017).

An example of a type 1 error would be when a researcher states their findings as significant when they happened by chance. An example of a type II error would be a researcher states there is no significant effect, when in fact, there is.

6) Explain why statistical significance does not tell us the magnitude of the relationship between the independent variable and the dependent variable.

Statistical significance will tell the researcher what the probability is that the relationship between the independent and dependent variables is due to random chance. When statistical significance is tested, the researcher will receive a probability between 0 and 1.0 (+ or-) which will tell them if the relationship was caused by chance (Rubin & Babbie, 2017).

REFERENCES

Rubin, A., & Babbie, E. R. (2017). Empowerment Series: Research Methods for Social Work (9th ed.). Cengage Learning.

Edited by Megan Gleicher on Jul 24 at 1:12pm