Data Handling and Analysis

Every time you bake a cake, you probably break down the process into different steps. You may pick a recipe, go to the shops, arrange the ingredients, and follow the recipe. Data handling is somewhat similar. Researchers follow a logical order to get the best quality end product possible.

Data handling is vital in any psychological study and directly impacts the quality of the results obtained. Let's take a closer look at data handling and analysis.

• The explanation will start by exploring data analysis in psychology, covering its uses and importance in research.
• Next, we will delve into the steps of data handling.
• Then we will review data handling and quantitative analysis together.
• Then, we will explore how data handling and analysis in research is carried out.
• And lastly, we will explore a data handling and analysis example in a research setting.

Data Analysis Psychology

Once data has been collected, there are several things that the researchers need to do, and one of these is data handling.

The steps that researchers take during data analysis are important because they can affect the validity and reliability of the findings. For instance, if the researcher found something unexpected and chose to ignore the variables they were initially interested in, the study will no longer be investigating what it intends to. Thus, lowering the validity of the study.

Moreover, if researchers don't use standardised procedures to analyse findings, this can lower the reliability of the study.

Steps of Data Handling

Now that we have learned the importance of data handling, let's uncover what steps data handling includes. There are six steps in data handling, which are:

Before a researcher analyses their research or even handles it, the researcher should have a clear and carefully plan on the direction of their research. For instance, before data is collected, the researcher should know the source of where they will collect the information.

Researchers can use primary or secondary data in their research.

Data analysis is a step that follows after a researcher has handled data.

In research methods, two types of data are collected: qualitative and quantitative. Qualitative data is processed slightly differently from quantitative data. Let's take a look at data handling in quantitative analysis.

Quantitative Analysis and Data Handling

Quantitative data is data that is expressed in numerical form. Quantitative analysis usually involves using a mathematical approach and statistics to identify whether the findings support or disprove hypotheses.

When carrying out statistical analyses, there are essential data handling steps that the researcher needs to consider. The reason is that statistical tests have certain conditions.

One of these conditions is identifying the level of measurement. There are four levels of measurement which essentially distinguish the different characteristics of variables. It is essential to understand the characteristics of variables because these will hint at which statistical analyses could be done and which could not.

Data analysis in psychology

As you learned, data analysis is the process in which statistical techniques are applied to find patterns within a sample. The data analysis in a study usually follows two steps. First, descriptive statistics are inspected, and then statistical analyses are performed.

As the name suggests, descriptive statistics describe the data's characteristics, and the two main types of descriptive statistical tests used are the measures of central tendency and measures of dispersion.

The measures of central tendency tests are used to calculate averages, and the three main types of tests are:

• Mean: the average score of the data set. It is calculated by adding all scores up and dividing them by the number of scores.
• Median: the central score in a given data set. It is calculated by putting all scores in order and picking the one in the middle.
• Mode: the most common score. It is calculated by looking at the scores and identifying which number has the highest frequency.

The measures of dispersion tests are used to measure the spread/ variance of the data. The common computations calculated are:

• Range: this is the difference between the lowest and the highest score.
• Standard deviation: the average distance of each score from the mean.

Data handling and analysis in research

Generally, descriptive statistics involve presenting the data. This presentation of data is usually done using graphs. The following are types of graphs:

Tables are used to show contrasts between a few sets of data. For instance, the table above shows the difference between control and drug conditions according to mean and standard deviation measurements.

Bar charts show the results of different conditions (or variables) using bars of different heights. Bar charts illustrate the differences between groups and make identifying trends and patterns easier.

Fig 1. Bar chart example

Scattergrams are also incorporated in correlational research. From scatterplots, researchers can identify the direction and strength of the relationship between two variables.

Fig 2. Scatterplot example.

Before inferential tests are conducted, researchers usually run descriptive analyses. Following this, the distribution of the data is analysed.

The type of distribution found will affect what statistical analyses can do later.

There are two types of distribution:

• Normal distribution forms a bell-shaped curve, as most data points are clustered towards the central values, while fewer data points are at the extremes.

• Skewed distributions, the data set can be positive (right) or negative (left).

Statistical Analysis

The statistical analyses employed in psychology research use inferential statistics to identify if the data supports or negates their hypothesis.

Sometimes we may find differences, but these may not be significant. The accepted level of probability in psychology is 0.05 (5%). Now, what exactly does this mean in research? Significance levels below this threshold suggest the observed change in the dependent variable is likely due to the manipulation of the independent variable. However, if it is above the threshold, the observed changes are likely due to chance.

The findings should not be accepted when research establishes differences, but the results are non-significant.

Data Handling and Analysis Example

Let's look into a real example of how data is handled and analysed in research.

In a study looking into the relationship between revision time and exam performance, researchers would first consider they will gather their data.

Initially, the researchers would carry out analyses concerning descriptions, e.g. the mean scores for revision time and exam performance. The findings may be reported as students in the current sample reported a mean revision time of around 6h (M = 5.78) and an average score of 78 points out of 100 in the exam (M = 78).

The two variables' distribution will be explored through a histogram. Depending on the visual inspection of the distribution, parametric or non-parametric tests would be performed. Let's imagine that the distribution of the two variables was normally distributed.

Researchers would then move on to data analysis, i.e. inferential statistics. To test the hypothesis, the researchers may run a correlational analysis. The Pearson correlation may be interpreted as the analysis shows a positive correlation between revision time and exam performance, r (20) = .78, p = .05.