Question:
What are common elements of statistical projects, presentations, etc?
?
2010-06-16 21:07:34 UTC
As part of my stats final, there's a project where we have to compile data, and portray the results statistically. My professor won't give out any information as to what's supposed to be covered, so does anyone know what the common elements are in statistical presentations. In class we've covered these topics: Descriptive Statistics, Probability, Discrete Distributions, Continuous Distributions, Sampling Distributions and Estimation, One-Sample Hypothesis Tests, Two-Sample Hypothesis Tests, Analysis of Variance, Bivariate Regression, Multiple Regression, and Chi-Square Tests (Non-Parametric). Within these topics does anyone know which are the most and least valuable for data representation? Thanks everyone!
Four answers:
2010-06-17 07:47:30 UTC
You generally start with a description of the research question and the data. You identify your dependent variable, and all the independent variables and any control variables used in the analysis. You describe the literature regarding previous studies in this area, and if you are building on or testing any of these studies, you want to say so.



Then, you tell how the data are gathered, what the sample was, and other information to allow researchers to get a handle on whether they think there may be problems in the study that are caused by deficiencies or misunderstandings in the data gathering process.



Next you describe the variables you have selected for your analysis, and what the theory is driving your selection of these variables. I think you put in your hypotheses (don't forget the nulls) at this point.



At this point you want to describe your analytical tools. Now this, of course, depends on what kind of variables you are using (binary, categorical, continuous) for both the dependent and independent variables.



Operationally, you do the analysis. Now, in reality, you might play around a lot with your data, and try various things, to see which gets you the best results. But in the presentation, you mention none of that fumbling around that everyone does. You act as if you knew it all from the beginning.



So, back to the presentation. You've got results. Display them. The form of the dispay must include your betas and their significance. If you like, you can put everything in a table. If you're good a graphics, you can display the data graphically. This is especially useful if you are doing a poster session. It looks cool, and gives you something to talk about. However, true statisticians look down their noses at such baubles. In the real world, if you do statistical analysis, you are going to be talking to ordinary people, as well as statisticians. I suggest you get in the habit of using graphical representation of your results, because people intuitively understand it better. Not a lot of people know what significance test are, nor the significance of them (sorry, couldn't resist -- it can be so dry discussing stats).



Finally, you tell your audience what the results mean. Did they answer your research question? Was your hypothesis proven, or did the null get proven? Well, of course, in the real world, hardly anyone presents papers where the null is proven. It's a shame, because it's just as important to know that black cats don't cause accidents as it would be to know they do. But in your case, since it's a student project, and you may not find anything in your data, you may well be presenting the null hypothesis. Well, it means something, so don't be ashamed.



I don't know if this helps, because I've spoken more generally about the goals of various parts of the paper, rather than telling you which tests and tools to use. Unfortunately, as I hope you understand, I can't tell you, without knowing what your research question and data are. But don't try giving me them, because I'm not gonna do it. One other thing. Perhaps at your school there is a statistical support service, or a data librarian? They can help with this, too.
?
2010-06-17 02:57:47 UTC
...that you not rely solely on advice from Askvillians who may not have studies statistics in many decades. Pick up several journals in your field at your university library--business, science, etc.--and see how statistical research, analysis, and conclusions are typically presented. I also recommend you take a look at some of the market and public opinion research studies available for free online. There typically are detailed reports which you also may find helpful; e.g., Pew, Gallup.



I find it very odd that your professor isn't eager to explain what his expectations are. I don't know if you study alone or in study groups. You may wish to seek opinions from your fellow students or teaching assistants. It would be a shame for us to make the wrong guess and have you work hard in a direction different from that your professor wishes you to take.



Good luck with this. If you can pick a subject of interest to you, this could be an interesting project.
2010-06-16 22:52:54 UTC
UNCC Final Exam

Elements of Statistics I Common Final Exam

http://www.math.uncc.edu/finals/stat122x/fall98/multichoice.pdf



Anova

Binomial distribution

Bivariate

Center

Chi-square distribution

Confidence interval estimation

Confidence intervals for p, a proportion

Confidence intervals for r

Confidence intervals for the mean

Data

Degrees of freedom

Descriptive statistics

Descriptives

Errors and residuals in statistics (type i and type ii)

Experimental design

Frequency distribution table

Histogram

Inferential statistics

Mean

Median

Mode

Multivariate

Normal distribution

Parameters

Probability

Probability distribution

Random variable

Regression

Regression toward the mean

Sampling

Sampling bias

Sampling distribution

Shape

Spread

Standard deviation

Statistical survey

T-distribution

Univariate

Variance





APA Statistical Abbreviations

• d -- effect size (Cohen)

• g -- effect size (Hedge)

• LR -- Likelihood ratio

• M -- mean

• Mdn -- median

• N -- total sample size

• n -- subsample size

• rs -- Spearman rho

• SD -- standard deviation

• SEM -- standard error of the mean

• SEM -- structural equation modeling (no italics)
?
2016-04-12 12:31:07 UTC
Sodium is a soft, silvery white, highly reactive element and is a member of the alkali metals within "group 1" (formerly known as ‘group IA’). It has only one stable isotope, 23Na. Sodium was first isolated by Sir Humphry Davy in 1807 by passing an electric current through molten sodium hydroxide. Sodium quickly oxidizes in air and is violently reactive with water, so it must be stored in an inert medium, such as kerosene. Sodium is present in great quantities in the earth's oceans as sodium chloride (common salt). It is also a component of many minerals, and it is an essential element for animal life. As such, it is classified as a “dietary inorganic macro-mineral.”


This content was originally posted on Y! Answers, a Q&A website that shut down in 2021.
Continue reading on narkive:
Loading...