Understanding the Data and Results
Information collected by researchers and clinicians about how well a patient is doing is called data. Quantitative data, which use numbers, can be gathered in several ways. Rating scales, such as a pain scale (0 to 10); graded performance, such as manual muscle testing (0 to 5); or number of errors (0 to 10) on a balance test are ways of measuring an aspect of someone’s health or performance. Qualitative data are word based and not represented by a number. Qualitative research considers individual experiences and beliefs to increase understanding about a topic. This type of data is not analyzed statistically but is sometimes included in a study to help explain why patients did or did not do as expected.
Which Exercise Equipment Is Best For Weight Loss Photo Gallery
Four types of data are used statistically. The first is nominal. These data categorize participants into groups. The groups are discrete, meaning that each participant can only belong to one group. Questions or tests that have yes or no answers produce nominal data. For example, the results of an x-ray are positive or negative. Ordinal data are the second type. This is a way of ranking a set of outcomes based on frequency, preference, or strength. Think of this as a team’s seating in a tournament. Teams are ranked according to their record. The number one team may be way better than the number 2 team, but the team in third is just a little below second place. The key to ordinal data is that the distances between the numbers are not equal. For example, an athletic trainer might be asked to rank a list of modalities in order of most used to least used. The number assigned to the modality is the data. Although the numbers follow in successive order, they do not represent exactly how often they are used, and one cannot say that modality number 4 was used twice as much as modality number 2. The next type of data is called interval. Intervals are equally spaced numbers on a scale so that one is the same distance from 2 as 2 is from 3. Surveys that ask you to rate your level of agreement from 1 (strongly disagree) to 5 (strongly agree) are creating interval data. Ratio data are the last type and are similar to interval data. However, ratio data include zero, so that a score can be compared in multiples of another score. For example, one athlete could jump twice as high as another athlete 24 vs 12 inches. As with interval data, the distances between numbers are equal. Ratio and interval data are analyzed using parametric statistics. Nominal and ordinal data are used in nonparametric statistics.
Parametric statistics are based on the idea that data are normally distributed in a sample. This means that most people would score in the middle, and there will be roughly equal numbers of people above and below average. Figure 3-4 shows how a set of interval data would look if they were normally distributed. Because a normal distribution is assumed, parametric statistics examine the means and distances from the means (standard deviations [SDs]) of a particular measurement. Parametric statistics assess the probability of a piece of data being a certain distance from the mean and produce results that are used in comparisons.
Nonparametric data do not distribute normally because the choices were limited. If putting students in groups by grade in school, you would have a group of seventh graders, a group of eighth graders, and a group of ninth graders. There would be no one between the 3 distinct grades (Figure 3-5). Here, finding the average or mean grade does not tell you much about the students. Knowing how many were in each group (frequency) is more important. Nonparametric data use the frequency of answers to look for patterns and to find differences between groups.
Maybe You Like Them Too
- THOSE BROWS! THAT HAIR! THOSE LIPS! THAT SKIN!
- The go-to For Mothers-of-The Bride Seeking Style and Comfort
- Deliciously Healthy NUTS FOR DOUGHNUTS
- 4 Best Workout Leggings and Tights
- HOW TO CREATE THE EXCITEMENT OF AN AFFAIR IN YOUR RELATIONSHIP