Analysis Blog #2

The goal of this analysis is to recognize factors suspected of influencing patients’ perceptions of the quality of healthcare received. Physician-patient interaction is an important factor that influences health outcomes and it is expected that if interactions with healthcare providers are perceived by the patient as being positive, then patients will seek routine medical attention, potentially leading to healthier life outcomes. This analysis attempts to determine what factors affect the belief of Americans that they receive quality healthcare from their healthcare professionals.

Patient perception of quality of healthcare has many implications for treatment and recovery. Many factors are associated with perceptions of quality of care received, including ethnicity, levels of trust & satisfaction with healthcare providers, feelings that patients’ concerns are heard, validated and addressed and a general sense that healthcare providers care about the individual needs of the patients. Levels of trust between patient and healthcare providers have been correlated with improved outcomes. Patients feeling satisfied with the provider-patient interactions and trusting in their healthcare providers are more likely to participate in their healthcare and follow treatment instructions.

Differences in the physician-patient relationship exist between ethnic groups. In analyzing satisfaction and trust of physician style, researchers found significantly lower levels of trust among minority groups as compared to whites. Levels of participation in their own healthcare among ethnic groups vary as well. African Americans are less likely to participate as compared to whites. Feelings of assertiveness during interactions with healthcare providers are lower among certain cultures.

I anticipated that patients who perceive the interactions with their healthcare professionals will also perceive the quality of care received as high. The following research hypothesis was proposed:

H1: Patients reporting receiving positive medical experiences will rate with quality of healthcare received higher than those patients reporting less than positive medical experiences.

Using data from the 2012 Health Information National Trends Survey (HINTS), this project uses as the dependent variable respondent ratings on the quality of health care received in the past 12 months. Using SPSS, independent variable responses were compared to ratings of quality of healthcare received. Ratings are scored on either a 4-point or 5-point Likert scale, where 1 indicates the highest rating and either 4 or 5 indicates the lowest rating. The findings of this analysis are presented below. Independent variables, considered to be factors in patient perceptions of receiving quality healthcare, were selected based on prior research discussed above.

In general, how would you say your health is?

In general, how much do you trust information about health or medical topics from a doctor?
In the past 12 months, how often did you feel that you could rely on your doctors, nurses, or other health care professionals to take care of your health care needs?
In the past 12 months, how often did your health professional give you the chance to ask all the health-related questions you had?
In the past 12 months, how often did your health professional give the attention you needed to your feelings and emotions?
In the past 12 months, how often did your health professional involve you in decisions about your health care as much as you wanted?
In the past 12 months, how often did your health professional make sure you understood the things you needed to do to take care of your health?

As predicted, when patients perceived that interactions with healthcare professionals were positive, patients indicated that they received quality healthcare. Patients that were given opportunity to ask all their health related questions, had their feelings and emotions attended to, felt their needs were being addressed, were included in their health care planning, and understood the things they needed to do to take care of their health reported receiving quality care at higher rates. In contrast, a weaker relationship was found between trust in the information received from the doctor and quality of healthcare received in the last 12 months. Likewise, patient’s perceptions of their general health was not highly correlated with the quality of healthcare received in the last 12 months

I unexpectedly  discovered there were no real disparities in perceptions of the quality of healthcare received between gender and race/ethnicity groups as well education levels.

Future research should perform deeper analysis of the effect that bedside manner of healthcare professionals has on health outcomes. It is important research to take on because if healthcare professionals can identify what factors influence patients’ perceptions of interactions with their providers, it is possible that patient outcomes can be improved.

Reflective Blog

I was a bit nervous to take this giant, scary STATS class; its been so many years since i have a math course let a lone statistics, but I have actually enjoyed the learning that has taken place and the feeling of accomplishing something that once made me feel a little anxious. SOCY 508 has introduce me to so many statistical pieces relevant to research, pieces I once struggled with (even in my first few semesters in the program), especially methodology and data analysis sections in reach articles. While reading these sections in , I was always a bit confused  as to what tests were being performed, and what method of sampling.  But having taken this course, I felt more confident and competent in my other course this semester- I had a better understanding  that helped out while writing my research proposal.

All of the course assignments really helped to reinforce my understanding of statistics. I found the recorded lectures and slides to be extremely helpful in reinforcing the content from the book. I love to work in SPSS so that was my favorite as far as assignments go, but the chapter & literature exercises were beneficial as a method of applying the material. The blogs were another great feature- I think I found these helpful because they allowed me to surf the web looking for examples while causing me to serendipitously stumble upon other relevant pieces information.  I do not think there were any features that did not enhance my learning.

As I stated in my course intro blog, I use SPSS a lot in my current job, data collection and analysis it is actually my favorite aspect or my job, so I fell even more familiar with it and its functions. I was quite excited to learn how to build charts and graphs at time of analysis rather than taking an extra step to create something in Excel. I am least familiar with the techniques for analyzing correlation (i.e, in which situations to use them), I know the basics and I know this will come with further use and experience so I am not worried.

While I had considered doing a group analysis project, I did not know the best way to carry this out with many folks being off campus and working. I think it is possible to do group work using tele/videoconferencing if schedules work out but I do think real-time communication would need to take place at some point during collaborations.

I really do not know what could be changed as far as improving the learning process for future students. I think all aspects of the course were clearly designed to present us with content, allowed us opportunity to apply & synthesize the content, and assess our knowledge. Pretty sure you effectively hit all objectives on Bloom’s Taxonomy of Education Objectives. Great Job!

 

Positive and Negative Relationships

When analyzing relationships between two variables, we can tell from the values whether a positive relationship is found or whether the variables are negatively related. A positive relationship, indicates that the values change in the same direction; high values on one variable are associated with high values , and conversely, low values on one are associated with low values on the other. An example of this relationship would be the relationship between years of education and the salary. Another would be level of educational attainment and income.  A negative relationship implies that the values change in the opposite direction; high values on one variable are associated with low values on the other. An example of this would be high unemployment rates in a community and low levels of social control within that community. Direction cannot be assigned when both variables in a table are dichotomous because directionality of the relationship can only be assigned to ordinal or interval-ratio, dichotomous variables are nominal.

This internet resource  is pretty straightforward in describing possible relationships between variables. The graphics and videos add to my understanding by providing a visual depiction of the directionality of relationships rather than just reviewing the numbers in a table as in the book. One can clearly see how the variable values either go up or down together or are inversely related.

 

 

 

Confidence Intervals

An increase in sample size is linked with an increase in precision of the confidence interval; the larger the sample size the more precise the interval becomes. As the confidence level goes down (e.g., from 99% to 95%) the confidence interval becomes more precise, or narrower in width. For a large sample size like the 2012 Health Information National Trends Survey (HINTS) (n= 1,500), it is acceptable to use a 95% confidence level. For smaller sample sizes, as in the case with a mean score on a class exam, it makes sense to use a 99% or 99.9% confidence level.

The image below is an example of a distribution. You should interpret from the image that the researcher can be 95% confident that the specified interval contains the true population mean.

norm2

 

 

 

 

 

 

 

 

 

 

 

Discuss and give examples of the types of situations in which an analyst would want to use a 95% confidence interval for estimation. Do the same for 99% and 99.9% confidence intervals. Locate and describe at least one internet resource that explains and illustrates the concepts of confidence levels and/or confidence intervals. What do you think it adds to the description in your textbook?

Analysis Blog #1

For my analysis I chose to examine Factors influencing perception of quality of healthcare received: What factors affect the belief of Americans that they receive quality healthcare? Using data are from the 2012 Health Information National Trends Survey (HINTS), this project uses as the dependent variable respondent ratings on the quality of health care received in the past 12 months. Independent variables, considered to be factors in patient perceptions of receiving quality healthcare, were selected based on prior research. The independent variables selected are as follows:

In general, how would you say your health is?

In general, how much do you trust information about health or medical topics from a doctor?

In the past 12 months, how often did you feel that you could rely on your doctors, nurses, or other healthcare professionals to take care of your health care needs?

In the past 12 months, how often did your health professional give you the chance to ask all the health-related questions you had?

In the past 12 months, how often did your health professional give the attention you needed to your feelings and emotions?

In the past 12 months, how often did your health professional involve you in decisions about your health care as much as you wanted?

In the past 12 months, how often did your health professional make sure you understood the things you needed to do to take care of your health?

 

All variables were measured on a 5-point Likert scale, excellent, very good, good, fair and poor.

As predicted, when patients perceived that interactions with healthcare professionals were positive, patients indicated that they received quality healthcare. Patients that trusted their healthcare provider, felt their needs were being addressed, and were included in their health care planning, reported receiving quality care at higher rates. As indicated in the graphs below, the greatest disparities in perceptions of the quality of healthcare received were seen between gender and race/ethnicity groups.

gender

 

 

Untitled

Sampling Distribution

An important thing to know about a sampling distribution that will help many understand the process is that it is not the distribution of a solitary sample, it is actually the process of selecting all possible subsets from a population and calculating a selected statistic, such as the mean or proportion, and using that information to make inferences about the population.

 

statistical-distributions-34-638

 

Sampling is the act of pulling all possible samples from the population and calculating the selected statistic and then putting it back in the “bucket”, and pulling another sample and calculating the mean and again putting that sample back in the bucket. This is called sampling because it is a process that is repeated. When you take each statistic, for instance the mean, from each sampling, and plot it on graph, theoretically each mean of all samples pulled will be close, although not exact, to the observed parameter of the population. The plotting will resemble the bell curve (see left).

 

 

 

mu

It is not realistic that researchers will be able to pull all possible combinations of samples, as this would be expensive and time intensive, but theoretically the information they gain from the samples of the population they do investigate will be similar to one another and cluster around a particular value, and this value can be interpreted as the same as the population value (see right).  These results are then considered to be generalizable to the population.

 

 

 

 

I think the below video is a good depiction of what the textbook attempts to represent in Figure 7.5 on page 223.  For some reason that graphic was hard for me to interpret-I think this video, and the graphics above add a little visual clarity to what the textbook tries to explain

 

 

 

The concept of a sampling distribution is often the most difficult concept in introductory statistics for students to grasp. Locate and present at least one internet resource that explains and illustrates the concept. What do you think it adds to the description in your textbook?

 

Grading Curve

normalcurve3

A normal distribution, or bell-shaped curve, is a theoretical model used to evaluate an empirical distribution of scores. Ideally, distributions should closely resemble the normal curve, with the median, mode and mean coinciding with each other at the exact middle of the distribution, and the scores should gradually decrease from the mean on each side of the middle, with approximately half of the scores falling below the mean and approximately half of the scores falling above the mean.

 

 

asymmetric_distribution_01

A test in which 75 percent of the scores fell below the mean, leads me to conclude that the scores are positively skewed and therefore not evenly distributed. In this case the mean is pulled in the direction of the high outliers above the mean, creating an inflated mean score.

 

 

The instructor should calculate the mean score and standard deviation, and adjust the distribution to fit the new normal curve so that 99% of the scores are within three standard deviations of the mean.

 

 

 

 

 

 

 

Imagine that you recently took a statistics exam and your instructor just returned your graded exam. The instructor announces that 75 percent of students scored below the mean. How do you reconcile this with the fact that, in a normal distribution, half the scores should fall below the mean and half of the scores should fall above the mean? What descriptive statistics do you think instructors should use to evaluate whether or not to curve exam scores? Can you describe a method of curving that you think is statistically sound?

Central tendency

For each group, the mean income is higher than the median, indicating a positively skewed distribution where there are few extremely high values. When looking at individual groups, the largest difference between mean and median is within the white workers group; there is nearly a $21,000 ($20,651) difference, well above the other two groups as well as all workers.  Black workers see a difference of $15,026 between mean and median income and the difference among Latino workers is $16,700. I think white workers have the greatest absolute difference between mean and median because more white workers than black and Latino workers occupy high paying jobs, thus creating the largest difference among white workers.

In this case the mean is pulled in the direction of the extreme high incomes, creating an inflated mean income level. When interval-ratio variables are skewed such as this, it is best to choose either mode or median as the measure of central tendency.  For this reason, I would choose median as that is the measure of central tendency that represents the exact middle, alleviating the skewed effect of the extreme values on the high end.

 

central

 

 

Examine figure below from the chapter’s “A Closer Look 4.2, A Cautionary Note: Representing Income.” Overall, how would you describe the differences between the median and mean incomes for each group? Why do you think white workers have the greatest absolute difference between median and mean incomes? If you had to report only one of these measures of central tendency, which one would you choose and why?

 

 

Chart Critique

 

 

 

508 chart

This chart is located at Dept. of Ed. projects public schools will be ‘majority-minority’ this fall.

I picked this chart because I thought the presentation was a bit different than I am used to (sort of a time series bar chart in histogram format) and it portrayed information in a clear, concise manner. This chart is designed to depict the shift in public schools from majority white to majority non-white since 1997, as well as the projected shift through 2022. I was able to understand the data immediately without having to read the accompanying description.

In my opinion, the data represented is best displayed in this format as this design has the ability to display a lot of information in one concise image. It provides a great visual impression of the demographic trends in the public schools. The data could be easily displayed in tabular form but the bars help the differences really stand out. Although there is a lot of information, I think the hard and fast facts are easy to pick out by the use of different colors on the bar. Additionally, using the dotted line to encompass the total percentages for each variable, majority vs. minority, draws your eye right to the important numbers.

Although I feel this visual is a great depiction of the data, I think this chart would not be easily read by someone not familiar with reading charts and graphs. This chart appears to be designed for persons having at least a basic understanding of data of statistics, which not all people do. Another draw back is that the actual total percentages for each minority group is not provided, neither in the accompanying text nor the chart.

To improve this chart, I would suggest adding actual percentages for each minority group, possibly by adding a notation under the chart. I think adding it to the chart itself may be too much data but I do think the data is important and would allow for comparison of trends within the minority groups. I think the readability may be increased by displaying the data in the traditional bar chart style.

 

 

Locate a recent chart or graph from the mass media or an academic source, and critique it. What’s good about it? What’s bad about it? Do you think it could be improved? If so, how?

Samples and Populations

A population study involves examining the entire population that have a particular set of characteristics that are of interest to the researcher. Studies where the total population is examined involve a relatively small group with characteristics that are not very common, for example employees of a specific organization, or parishioners of the local church. A key feature of a population study is ensuring that EVERYONE in the target population is studied, thus lending itself to research studies involving small populations.  Population studies can become expensive as a result of the additional resources (e.g., funding, legwork, etc.) needed to conduct, therefore population studies are not feasible for all research questions, such as in instances when the target population would be very large. For example, it would be nearly impossible to survey all people who attend church all across America, or even all people who attend church in only one county in America.

An example of a research question where it would be advisable to study the entire population, as opposed to just a sample, is the effect of a professional development class on customer service performance of all 215 customer service representatives in XX Organization. The target population is all customer service representatives within that organization. A second example is studying the number of times the 500 parishoners of XX Church received spiritual counseling in a month and the effect of that counseling on their feelings of happiness. The target population is all parishioners at XX Church. Both of these examples use an entire population, rather than just a subset of the target population.

Using the populations described above, to move this research to a sample study by analyzing our variables in a subset of the target population, we could look at the effect of a professional development course on customer service performance of 75 customer service representatives chosen randomly from the entire population of the customer service representatives in XX Organization. Likewise, we could study the number of times a random sample of 250 parishioners at XX Church received spiritual counseling in the last three months and the effect that counseling on their feelings of happiness.

Two paths to finding out what you want to learn; it is important to know which one to use!

ramdom-sample 508

 


Are there any research questions or specific hypotheses where it would be advisable to study the entire population, as opposed to just a sample? If so, what characterizes these situations? Which types of research questions and hypotheses lend themselves to a study of the entire population? Which types do not and why?