PREVNEXTSUBMITMenuResourcesGlossaryTranscriptClear and return to MenuSearch ResultsFilterSlide TextTranscriptSearch in:Check to includeSearch...TermsDefinitionFINISHfinishdefinitionresourcessearchpauseplayreplaysubmitnextpreviousvolumeSkip Navigation. Hit enter to return to the slide.SUBMIT ALLsubmit allQuestion ListMore infoSend an emailCloseBioExit<html><textformat leading='0' tabstops='[0, 48, 96, 144]' blockindent='0' indent='0'>Welcome to part 2 of module 4: Basic Numeracy in Health Care: Numerical Measures, Distributions, and Data Patterns. In this part, we’ll look at how to define and distinguish between numerical and categorical data.</textformat></html><html><textformat leading='0' tabstops='[0, 48, 96, 144]' blockindent='0' indent='0'>In module 2, we focused on counts and how to review (or disaggregate) counts by category, such as age or sex of the patient. Remember how Sister Zingy summarised the diarrhoeal disease outbreak by using frequency tables showing the number of persons found with diarrhoeal disease, categorised by age group and sex? So far in module 4, we have covered quantitative measures used in health, such as measures of weight, blood pressure, or haemoglobin. Now we’re going to re-introduce variables and discuss the different types of data that we can collect and use to describe or characterise humans and human health.</textformat></html><html><textformat leading='0' tabstops='[0, 48, 96, 144]' blockindent='0' indent='0'>In module 2, we defined a variable as a quantity or measure that can differ from person to person (or object to object). They’re called variables because the values of these characteristics can vary. </textformat></html><html><textformat leading='3' tabstops='[0, 48, 96, 144]' blockindent='0' indent='0'>Here’s a table showing examples of variables.</textformat></html><html><textformat leading='0' tabstops='[0, 48, 96, 144]' blockindent='0' indent='0'>We’ve seen the many different types of data Sister Zingy can use to characterise her patients, including Lebo at the ANC clinic or Andiswa at the ART clinic. Now let’s talk about the many types of data she has been dealing with: sex of patient, year, age, CD4 count, weight, and type of ART regimen. How Sister Zingy summarises her clinic’s data will depend on the type of data she is working with.</textformat></html><html><textformat leading='0' tabstops='[0, 48, 96, 144]' blockindent='0' indent='0'>In part 1, we discussed examples of quantitative variables: weight, haemoglobin, and blood pressure. In this section, we’ll discuss categorical variables, such as sex, which place an individual (or observation) into one of several groups or categories. A categorical variable can also be referred to as a ‘factor’ or ‘qualitative variable.’</textformat></html><html><textformat leading='0' tabstops='[0, 48, 96, 144]' blockindent='0' indent='0'>As we saw in module 2, categorical variables sometimes have no natural order or sequence that determines how to view or record them. One such example is patient sex. When summarising a count of patients by sex, we may choose to present the number of males first, or the number of females first. Either would be considered correct, as there’s no obvious natural order. Another example is marital status—there is no specific order for “single,” “married”, or “widowed or divorced”.</textformat></html><html><textformat leading='0' tabstops='[0, 48, 96, 144]' blockindent='0' indent='0'>Some categorical variables do have a clear natural order; we find many examples in the health care setting. Severity of infection, for example, can be categorized as mild, moderate, or severe. The WHO stages of HIV constitute another example of a categorical variable with a natural sequence, from Stage I through Stage IV. </textformat><textformat leading='0' tabstops='[0, 48, 96, 144]' blockindent='0' indent='0'>Though they can have some similarities, categorical variables are different from numerical measurements, such as we use for weight. For example, ‘moderate’ is clearly more than ‘mild’ and less than ‘severe’, but we can’t measure the exact difference between ‘moderate’ and ‘mild’. In other words, unlike true numbers, you cannot meaningfully calculate the distance between these values.</textformat></html><html><textformat leading='0' tabstops='[0, 48, 96, 144]' blockindent='0' indent='0'>As we have seen, a quantitative, or numerical, variable takes number values for which maths operations—addition, subtraction, multiplication, and division—make sense. The values of a quantitative variable are usually recorded as a unit of measurement, such as seconds or kilograms. </textformat></html><html><textformat leading='0' tabstops='[0, 48, 96, 144]' blockindent='0' indent='0'>There are some quantitative variables where only measurement in whole numbers make sense. These are called discrete variables and they are used in cases where we’re counting something that can’t, because of its nature, be divided. An obvious example of this would be people, because we can’t have fractions of people! A family or household, for example, can have 5 or 6 members, but not 5.5 members. Another example is the number of patient visits to a clinic. It’s possible for a clinic to have 40 visits in a day, but not 40.5.</textformat></html><html><textformat leading='0' tabstops='[0, 48, 96, 144]' blockindent='0' indent='0'>Other quantitative variables can have any value, including fractional or decimal values—and can even include infinity. These are called continuous variables. Weight is one example. Weight can be rounded to the nearest kilogram (75 kg), or it can be written more precisely, by including decimal places (74.87 kg). This data is considered to be continuous because it can be any number or fractional amount. Other examples are laboratory counts (like bilirubin or haematocrit levels), blood pressure, body temperature, and many other physiological measures.</textformat></html><html><textformat leading='0' tabstops='[0, 48, 96, 144]' blockindent='0' indent='0'>In this part, we looked at the differences between quantitative and categorical data. It’s important to be aware of the kinds of variables we are using when we work with data, because they determine how we analyse and summarise that data. </textformat><textformat leading='0' tabstops='[0, 48, 96, 144]' blockindent='0' indent='0'>Now let’s move on to part 3, where we’ll talk about how to display summaries of quantitative measures through frequency tables, histograms, and trend and scatter plots.</textformat></html><html><textformat leading='0' tabstops='[0, 48, 96]' blockindent='0' indent='0'>Categorical variables have values that describe a quality or characteristic. They use categories rather than numbers. For example, the categorical variable “gender” could contain the categories “Male”, “Female”, and/or “Other.”</textformat><textformat leading='0' tabstops='[0, 48, 96]' blockindent='0' indent='0'></textformat><textformat leading='0' tabstops='[0, 48, 96]' blockindent='0' indent='0'>Categorical variables can be ordinal or nominal. If observations can take a value that can be logically ordered or ranked it is an ordinal variable (e.g., strongly agree, agree, disagree, strongly disagree). If observations cannot be organised in a logical sequence it is a nominal variable, e.g., gender, profession, college attended, political affiliation and disability status. </textformat></html><html><textformat leading='0' tabstops='[0, 48, 96]' blockindent='0' indent='0'>The most commonly investigated characteristic of a collection of data (or dataset) is its centre, or the point around which the observations tend to cluster. Measures of central tendency measure the middle or centre of a distribution of data. Also called centre and central location. Examples are mean, median, mode.</textformat></html><html><textformat leading='0' tabstops='[0, 48, 96]' blockindent='0' indent='0'>The sum of the previous relative frequencies.</textformat></html><html><textformat leading='0' tabstops='[0, 48, 96]' blockindent='0' indent='0'>Decimals are numbers written in a special way to show the exact numerical value of a fraction. Numbers to the left of the dot show the part that is equal to 1 or more, and numbers to the right of the dot show the part that is less than one. </textformat><textformat leading='0' tabstops='[0, 48, 96]' blockindent='0' indent='0'></textformat><textformat leading='0' tabstops='[0, 48, 96]' blockindent='0' indent='0'>For example, the fraction ¼ is actually 1 divided by 4 which written as a decimal = 0.25. The fraction 1¼ would be written as 1.25. </textformat></html><html><textformat leading='0' tabstops='[0, 48, 96]' blockindent='0' indent='0'>To separate something into its component parts. </textformat></html><html><textformat leading='0' tabstops='[0, 48, 96]' blockindent='0' indent='0'>The way that something is divided or spread out.</textformat></html><html><textformat leading='0' tabstops='[0, 48, 96]' blockindent='0' indent='0'>A graphical display where the data is grouped into ranges (such as "40 to 49", "50 to 59", etc.), and then plotted as bars (without space in between). These are typically used for moderate to large data sets.</textformat></html><html><textformat leading='0' tabstops='[0, 48, 96]' blockindent='0' indent='0'>Refers to the difference between the value of a dataset at the 75th percentile and the 25th percentile. The IQR is often presented with the median to summarise central tendency and spread.</textformat></html><html><textformat leading='0' tabstops='[0, 48, 96]' blockindent='0' indent='0'>When looking at the distribution of a dataset, this means that the left side (or tail) extends much farther out than the right side and the mean is on the left side. Because the left side is the negative side of the peak, it is also referred to as “negatively skewed."</textformat></html><html><textformat leading='0' tabstops='[0, 48, 96]' blockindent='0' indent='0'>The average value of a set of numbers.</textformat></html><html><textformat leading='0' tabstops='[0, 48, 96]' blockindent='0' indent='0'>The middle value in a series of values arranged from smallest to largest. The median is another measurement of central tendency but it is not as sensitive to extreme values as the mean because it takes into consideration the ordering and relative magnitude of the values.</textformat></html><html><textformat leading='0' tabstops='[0, 48, 96]' blockindent='0' indent='0'>The most frequently occurring value in a dataset. The most common value. A histogram with a single peak has one mode. A histogram with 2 peaks has 2 modes. We call this type of histogram “bimodal.”</textformat></html><html><textformat leading='0' tabstops='[0, 48, 96]' blockindent='0' indent='0'>A special type of distribution which follows a bell shaped curve. The normal distribution is an example of a symmetric distribution and is not skewed to the left or the right.</textformat></html><html><textformat leading='0' tabstops='[0, 48, 96]' blockindent='0' indent='0'>Measurements that have a number value rather than a categorical one, like male or female. Examples include weight, haemoglobin, and blood pressure.</textformat></html><html><textformat leading='0' tabstops='[0, 48, 96]' blockindent='0' indent='0'>Values that "lie outside” of the majority of values. Outliers are not like most of the other values in a data set because they are much smaller or larger than most of the other values in the dataset. They are extreme values.</textformat></html><html><textformat leading='0' tabstops='[0, 48, 96]' blockindent='0' indent='0'>One of 100 equal parts that a dataset can be divided into in order to rank them.</textformat></html><html><textformat leading='0' tabstops='[0, 48, 96]' blockindent='0' indent='0'>A variable that can have any numerical value, even infinity. It is “continuous” because its value does not have to be a whole number – it can be any fraction in between. One example of this is weight. Weight can be rounded to the nearest kilogramme (75 kg), or it can be written in a more exact manner with decimal places (74.877 kg).</textformat></html><html><textformat leading='0' tabstops='[0, 48, 96]' blockindent='0' indent='0'>A variable that is numerical in value but must be a whole number. These are used in cases where we are counting something that cannot be less than one, such as number of people, equipment, or services provided.</textformat></html><html><textformat leading='0' tabstops='[0, 48, 96]' blockindent='0' indent='0'>The highest and lowest possible amounts in a series of numbers. For example, the range might be from 3 to 6.</textformat></html><html><textformat leading='0' tabstops='[0, 48, 96]' blockindent='0' indent='0'>How often something happens divided by all outcomes. For example, if a football team plays 12 games total and wins 9, the relative frequency equals 9/12 or 75%.</textformat></html><html><textformat leading='0' tabstops='[0, 48, 96]' blockindent='0' indent='0'>When looking at the distribution of a dataset, this means that the right side (or tail) extends much farther out than the left side and that the mean is on the right side. Because the right side is the positive side of the peak, it is also referred to as “positively skewed."</textformat></html><html><textformat leading='0' tabstops='[0, 48, 96]' blockindent='0' indent='0'>A graph which shows one dot for each measurement. Scatter plots can help to show the relationship between two variables. </textformat></html><html><textformat leading='0' tabstops='[0, 48, 96]' blockindent='0' indent='0'>The form or outline of an object. When referring to the shape of a dataset or distribution, it can be characterized by the number or modes (peaks) as well as symmetry. </textformat></html><html><textformat leading='0' tabstops='[0, 48, 96]' blockindent='0' indent='0'>The shape is not symmetrical. There are either a few observations much smaller than the rest, or a few that are much larger. The distribution is said to have a long “tail” and can skew left (negative) or right (positive). </textformat></html><html><textformat leading='0' tabstops='[0, 48, 96]' blockindent='0' indent='0'>When referring to the spread of a distribution or dataset, you are measuring how much variation there is, the range of data, and whether there are any outliers. </textformat><textformat leading='0' tabstops='[0, 48, 96]' blockindent='0' indent='0'></textformat><textformat leading='0' tabstops='[0, 48, 96]' blockindent='0' indent='0'>The shape shows us if our data has a tight spread, with all of the data points stacked closely together, or a wide spread. Measures include range, variance/standard deviation, and interquartile range. </textformat></html><html><textformat leading='0' tabstops='[0, 48, 96]' blockindent='0' indent='0'>Measures how far values deviate from the mean; in calculation, it is the square root of the variance.</textformat></html><html><textformat leading='0' tabstops='[0, 48, 96]' blockindent='0' indent='0'>A number that represents a piece of information such as information about how often something is done, how common something is, etc.</textformat></html><html><textformat leading='0' tabstops='[0, 48, 96]' blockindent='0' indent='0'>When the shape of a distribution or dataset is symmetrical, the mean is exactly in the middle and the left and right sides are (approximately) equal to each other.</textformat></html><html><textformat leading='0' tabstops='[0, 48, 96]' blockindent='0' indent='0'>Any characteristic or attribute that can be measured and can have different values. Height, weight, sex, age, and CD4 are all examples of variables.</textformat></html><html><textformat leading='0' tabstops='[0, 48, 96]' blockindent='0' indent='0'>The amount of difference or change from the mean.</textformat></html>