Types of Data

Quantitative data

• Data in the form of numbers (quantity)

• The recording of variables collects quantitative data eg reaction times using a stopwatch, a rating emotion out of 7 using a Likert scale or a tally of the number of times someone performs an operationalised behavioural category in an observation

Quantitative data → Descriptive statistics

Descriptive statistics (averages and ranges) summarise quantitative data, and these descriptive statistics are then displayed on tables and graphs

Qualitative data

• Data in the form of words

• These words are descriptions of behaviour, thoughts and/or feelings

Qualitative data → Content analysis

Content analysis converts large amounts of qualitative data into quantitative data. To turn observations and interviews into qualitative data behavioural categories can be created and then tallied

When to use quantitative and qualitative data

• Quantitative data is used in experimental and observational research

• Qualitative data is used in case studies, open question interviews and questionnaires

Studies that use both quantitative and qualitative data

Studies can collect a combination of both quantitative and qualitative techniques in research. If both methods agree, this increases credibility (methodological triangulation)

AO3 - Advantages of quantitative data

• Objectively measured, reducing the likelihood of bias. This increases scientific credibility

• Descriptive statistics allows quantitative data to be summarised and then displayed on graphs, charts and tables

• Quantitative data tends to be more reliable because of the limited number of responses, there is a higher chance of getting the same findings if the study is repeated

AO3 - Advantages of qualitative data

• Quantitative data is seen as rich in detail, this is because qualitative researchers often collect more information, and the use of open ended questions means participants are not limited in the responses they can give, meaning qualitative data has higher validity (eg participants can give the answer they want, not limited to yes/no or out of 7)

AO3 - Disadvantages of quantitative data

• The limited number of qualitative research responses results in data lacking depth and detail. Also, qualitative data collection can only focus on individual behaviours and what can be mathematically measured

AO3 - Disadvantages of qualitative data

• Qualitative data gathered by the researcher can be open to interpretation and potentially biased

• Due to the extensive range of data collected, it can be challenging to summarise

• As the questions that produce qualitative data are open ended, this tends to be more variable, reducing the reliability of qualitative research

What is primary data?

The researcher is responsible for generating the data, also known as ‘first hand’ or ‘original’ data. Primary data is created to answer the research question. Common ways to collect primary data are the researcher conducting experiments, observations, interviews, questionnaires and case studies

What is secondary data?

Also known as ‘second hand’ data, this is when researchers use information previously collected by a third party, such as another researcher or organisation. This secondary data was initially collected for a reason other than to answer the current research question. Examples of secondary data are government or business statistics and records or previously published studies

AO3 - Advantages of primary data

• Increased validity as the data is collected to answer the research question directly. The experiment or observation is designed to test the intended variable directly

• Increased validity as the researcher can control the data collection process carefully

AO3 - Advantages of secondary data

• Secondary data already exists and is often already analysed, this can dramatically reduce both the time needed to conduct research and the costs involved in conducting a study involving participants

AO3 - Disadvantages of primary data

• Collecting original data from participants is both time consuming for the researcher and potentially expensive. Costs include paying participants for their time and other researchers for their work. Setting up an experiment also includes paying for materials

AO3 - Disadvantages of secondary data

• Decreased validity as the data is not collected to answer the research question directly. The data may not be appropriate to answer the researcher’s research question

• Decreased validity as the researcher had no role in the data collection process, so cannot ensure that the data was collected free from bias or is the result of variables

What is a meta analysis?

A process that collects and combines the results of a range of previously published studies asking similar research questions. The data collected is compared and reviewed together, and part of this review can include statistically combining all the data to produce an overall effect size and conclusion

AO3 - Advantages of meta analysis

• The large sample size of meta analysis produces results that are more statistically powerful than studies with a small number of participants

• As meta analysis looks at the overall pattern of results across many studies, a small number of individual studies that are affected by bias or a lack of control will not change the overall pattern of results, making meta analysis more trustworthy than any individual study

• Studies testing the same variable in various contexts (such as across cultures) be compared, revealing unexpected relationships

AO3 - Disadvantages of meta analysis

• A meta analysis has all the weaknesses of secondary data, the researcher has no control over the quality of the data collected. Also, included studies are conducted to answer particular research questions, so may not be comparable

• Studies shoe that a statistically significant result are more likely to be published (so included in a meta analysis) while non significant results are unlikely to be submitted for publication (the file draw problem)

• The choice of which studies to include/exclude could be biased