Essay Instructions: Dear writer, below is an explanation of how the instructor wants this paper written. Basically using the different paragraphs further below please summarize them in one page paper. Thank you
Dear Sheila
What you need to do is examine different types of statistics/statistical tests as well as when and why these particular tests would be selected for use. In addition, you need to mention what the results indicate. So basically, you write he name of the test, define why it is used and what do we understand from the results. I think one page should be enough because the idea is to summarize the material. The 1-page study sheet should include the key concepts of the chapter: Focus on the basic concepts that are important for nurses to understand as they review research studies and do not go into two much details. Once again, go point by point. I,e:
•&????;Regression analysis: ……………………………………
•&????;Analysis of Variance (ANOVA) ………………………………….
•&????;Chi Square: …………………………………..
In the dotted line just write the summary.
Hope this helps.
To prepare:
• Review the information in your assigned chapter.
• As a group, develop a 1-page study sheet that includes the following:
o The key concepts of the chapter: Focus on the basic concepts that are important for nurses to understand as they review research studies.
o A description of the statistical methods covered in the chapter, what they measure, and under what circumstances they are used. Identify examples of how the statistical methods have been used in research studies.
o An explanation of the key statistical tests and how they measure significance (if applicable).
Note: This should be a collaborative effort, with each member of the group making contributions to the design and content of the study sheet. Use the Groups link on the left navigation bar to collaborate with your group. When you have developed your 1-page study sheet, select one member to post it to the Week 9 Discussion Forum so that the rest of your colleagues can access it.
Post on or before Day 5 your group’s study sheet. Discuss why it is important for nurses to understand the basics of these statistical methods.
Key Concepts for Research Studies
Basic Concepts for Nurses to Understand
Inferential statistics are based on the laws of probability and allow conclusions about a whole population from a sample. Probability samples that are chosen randomly using large samples ensure representative samples for a population. This is important for drawing conclusions from a sample. Sampling error and standard error of the mean (SEM) reflect that each sampling distribution has some error when estimating for the population. The smaller the SEM, the more accurate is the sampling distribution (Polit & Beck, 2012).
Statistical inference for the effectiveness of interventions on a group can be obtained using confidence intervals (CIs), point estimates and hypothesis testing. CIs are a range of values that a parameter has a probability of lying within and include estimate of error. Point estimates do not include estimate of error, so are unable to be assessed for accuracy. The null hypothesis always states that there is no relationship between groups because of the intervention. The alternative hypothesis states that there is a relationship because of the intervention. Accepting a null hypothesis in error is a Type I error. Rejecting the null hypothesis would mean a Type II error (Polit & Beck, 2012). CIs and hypothesis testing are reliable predictors of intervention effectiveness.
All studies should give the results for hypotheses used in the study and should address analyses that reflect internal validity. Each study should be evaluated for statistical analyses and design to assess appropriateness. Independent variables should be defined clearly, sample sizes should not be small and statistical analysis should be powerful to decrease risk of drawing the wrong conclusion about the research hypothesis.
References
Polit, D. F., & Beck, C. T. (2012). Nursing research: Generating and assessing evidence for nursing practice (9th ed.). Philadelphia, PA: Lippinkott Williams & W
Inferential Statistics
Description of Statistical Methods Covered and How They are Used
Probability Sampling: Used to obtain representative sampling-the preferred method for obtaining statistical validity with inferential statistics. However, even when random sampling is used, the results are rarely identical to population characteristics (Polit & Tatano Beck, 2012). The goal of the researcher is to avoid sampling error, obtaining sample values that are as true to population parameter as possible. The sampling distribution is a distribution of the mean, and is theoretical rather than actual, and forms the basis of inferential statistics (Polit & Tatano Beck, 2012).
Standard error of the mean (SEM): Indicates the degree of average error of the sample mean, with the smaller SEM indicating greater accuracy of the sample estimate, and is figured using the sample’s standard deviation and sample size. “Because the SEM is partly a function of sample size, we need only to increase sample size in order to increase the accuracy of our estimate” (Polit & Tatano Beck, 2012, p. 405).
Statistical inference: Consists of estimation of parameters and hypothesis testing.
Parameter estimation estimates a mean, proportion or mean difference between two groups, using either point estimation or interval estimation (Polit & Tatano Beck, 2012).
Point estimation calculates “a single descriptive statistic to estimate the population sample” (Polit & Tatano Beck, 2012, p. 406), and cannot objectively be used to convey information regarding margin of error. Interval estimation indicates “a range of values within which the parameter has a specified probability of lying” (Polit & Tatano Beck, 2012, p. 406). Researchers create a confidence interval (CI), with upper and lower limits around the estimate, called confidence limits.
Hypothesis testing: provides objective criteria for deciding if the hypothesis is supported by data. The Null hypothesis indicates there is no relationship between variables. Statistical hypothesis testing is a process of rejecting probability of study results reflecting a true population difference.
The use of both parameter estimation and hypothesis testing is used when evaluating evidence based practice studies, and many of the research journals are actually requiring that this be reported along with the research, noting its value to clinicians (Polit & Tatano Beck, 2012).
• To know the standard deviation (SD) of the sampling distribution allows the accuracy of a sample mean to be interpreted as shown in Figure 17.1.
• “An easier method for constructing 95% CIs around major risk indexes is to use the University of British Columbia’s ‘Clinical Significance Calculator’ on the Internet (http://spph.ubc.ca/sites/healthcare/files/calc/clinsig.html)” (Polit & Beck, 2012, p. 407)
• Figure 17.2 illustrates the outcomes of statistical decision making of Type I and Type II errors of null hypothesis
• Type I error is a false positive result (alpha=?)
• Type II error is a false negative result (beta= ?)
• Level of Significance (alpha or ?) is a method used to reduce the risk of Type I Errors in Null Hypothesis: acceptable values for ? are .05 and .01
• Power Analysis: is a method used to reduce the risk of Type II Errors of Null Hypothesis including the following four components to estimate the needed sample size:
1. the significance criterion, (?)=.05
2. the sample size, (N)
3. the effect size, (ES) example: Cohen’s d
4. power, or (1 ??" ?): conventional standard is .80
Polit, D. F., & Beck, C. T. (2012). Nursing research: Generating and assessing evidence for nursing practice (Laureate Education, Inc., custom ed.). Philadelphia, PA: Lippincott Williams & Wilkins.
Key Concepts for Research Studies
Explanation of the key statistical tests, and how they measure significance?
Every test statistics has a related theoretical allocation and with equivalent process for testing hypothesis. Researchers use test statistics to compare value of computed test statistics to value of critical limits (Polit & Beck, 2012). Critical limits that are above test statistic are referred to as being statistically significant (Polit & Beck, 2012). Significant is the outcome that is less likely related to chance.
Two broad classifications for statistical test are parametric estimation of parameters verses nonparametric which requires no estimation and less restrictive assumptions of variables. The preferred and powerful method for statistical testing is parametric. Nonparametric statistical testing also know as distribution free statistics is more favorably used when data cannot be construed in any manner, distribution is markedly non-normal, and in small sample size (Polit & Beck, 2012).
Since statistical test are rooted in comparisons designs style will vary depending on the types of groups being study for example Nominal level variable, ordinal, interval or ratio (Polit & Beck, 2012). Two standard statistical test used are t-test and analysis of variance ANOVA used when testing significance of difference among groups. Other types of nonparametric t-test and ANOVA are Mann-Whitney U test (two group), Wilcoxon sign-rank test (two group), Kruskal-Wallis (three group), Friedman tests (three group), Chi-square test (proportional difference), and Fisher’s exact test (small sample) (Polit & Beck, 2012).
Statistical test that use descriptive and inferential correlations are Pearson’s r (interval-level data), and subsets Spearman’s (ordinal data), Kendall’s tau (ordinal data), phi coefficient (nominal data), and Cramer’s V (nominal data) (Polit & Beck, 2012).
References
Polit, D. F., & Beck, C. T. (2012). Nursing research: Generating and assessing evidence for nursing practice (9th ed.). Philadelphia, PA: Lippincott Williams & W
?
