Research Methods Pt3 (Selecting Pp - Qualitative Data)

Research Methods Pt3 (Selecting Pp - Qualitative Data)

1. Selecting Participants

   • Random Sampling
     • Individual members are assigned an identifier, such as a number, and the required number of participants is then selected using random number tables or computer generated random numbers.
       • Unbiased Because no researcher bias Therefore increased validity
       • Pp variable should be spread across groups Because random allocation Therefore indicates C+E
       • Requires more work Because create list of target pop Therefore more time consuming and costly
       • Doesn’t guarantee a representative sample Because still may produce biased sample Therefore difficult to generalise
   • Opportunity Sampling
     • Selecting anyone who is available to take part from the given population until sufficient numbers have been obtained.
       • Requires less work Because use who’s available Therefore less time-consuming and costly
       • Good for pilot study Because quick to test procedures therefore improves main study
       • Unlikely to be representative Because may produce bias sample Therefore difficult to generalise
       • Research bias Because researcher chooses Pp to fit research Therefore reduces validity
   • Self-selected/Volunteer Sampling
     • Pp become part of research because they volunteer when asked.
       • Participant bias Because Pp will be highly motivated and perform better Therefore reduces validity
       • Unlikely to be representative of the target pop Because may produce bias sample Therefore difficult to generalise
       • Requires less work Because researcher not required to spend time gathering Pp Therefore less time consuming and costly
       • Pp want to take part of research Because self-selected Therefore less drop out
   • Systematic Sampling
     • An organised form of selection using ready prepared lists of the target population.
       • Requires less work Because using prepared lists of the target population Therefore less time consuming and expensive
       • Doesn’t guarantee a representative sample Because still may produce biased sample Therefore difficult to generalise
   • Stratified Sampling
     • Used when the population has important sub-groups (strata). Random sample are then taken from each stratum in the same proportions, as they exist in the population.
       • More representative sample Because sub-groups Therefore more generalisable
       • No researcher bias Because random sample of sub groups Therefore increased validity
       • Requires most work Because create list of target population of each strata Therefore most time-consuming and expensive

2. The Ethics of Research

   • Debriefing
     • A post research interview designed to inform participants about the true nature of the study. The aim is to restore participants to that state they were in before the start of the study.
     • It also gives participants the opportunity to ask any questions, receive follow-up help (if needed) and give feedback about the study to the researcher.
   • Ethical Committee
     • A group of people within a research institution that must approve the study before it begins. It normally consists of professionals from within the area, who consider what ethical issues could arise, how to deal with them and weighs up the cost-benefit issues.
   • Presumptive Consent
     • A method to deal with lack of consent or deception.
     • A group of people similar to the group of participants are asked if they would agree to take part in the study, if the group of people agree, then it is presumed that the participants would also.
   • Right to Withdraw
     • Participants have the right to withdraw from a study if they feel uncomfortable in any way.
     • Participants should be reminded of this throughout the study.
   • Confidentiality
     • All participants have the right to have their personal information protected. This means that any information collected needs to be stored securely and destroyed as soon as the data is no longer required.

3. Reliability and Validity

   • Reliability looks at the measure of consistency.
     • Internal Reliability
       • This is how consistent a method is within itself e.g. If the same test is given to all participants then it is reliable, if different tests are given out then it isn’t. If materials (instructions, debriefs etc.) are not standardised, different tests are given to participants or if observers produce different results for the same situation then internal reliability can be affected.
       • To test internal reliability you can use the split half technique. This is where the results for the study are halved and analysed separately, before being compared again. If the results are the same or similar, then they have not been affected by internal reliability issues.
     • External Reliability
       • This is how consistent a method measures over time when repeated e.g. If a participant is rated as extrovert by a personality test on one day, then they should be rated as extrovert if repeated a week later. Different materials, tests, researchers scoring data differently on separate occasions and if participants change their mind or level of attention throughout the study can all affect the external reliability of the research.
       • To test external reliability you can use the split-half technique. This is where the results conducted on one occasion are correlated with the results taken on another occasion. If the results have a high correlation then they have not been affected by external reliability.
   • Validity looks at the extent to which a study and its findings are legitimate and true.
     • Internal Validity
       • This is concerned with if a piece of research measured what it was supposed to measure e.g. An IQ test should ask questions about perception and arithmetic and not personality. Factors that can affect internal validity include if a participant answers dishonestly (social desirability), participants have not been made aware of confidentiality guidelines or if a test is not assessing what it was supposed to investigate.
       • To test internal validity you can use content validity or concurrent validity. Content validity looks at the content of the test to see if it measures what it is supposed to measure and concurrent validity compares the new method with an established test.
     • External Validity
       • This looks at if the results can be applied to different environments or different people e.g. If an experiment was conducted in artificial settings it would have low external validity, but if it were conducted in natural settings it has high external validity. Factors that can affect external validity include unrepresentative samples, artificial settings, research only being conducted in one culture or older research being applied to today.
       • To test external validity you can repeat the experiment using different participants, cultures, environments or at a different time.

4. Qualitative Data

   • Qualitative data is normally data in the form of words. It is rich data and therefore full of quality, deep feelings and emotions. However, it is also the least scientific type of data. It can be taken from interview scripts, field notes, diaries, questionnaires, observation notes and audio recordings.
   • Qualitative data gives very detailed accounts of people’s opinion and emotions because participants are not restricted on how they answer therefore the results are rich and full of meaning.
   • Qualitative analysis can be used to analyse anything because it takes into account all aspects of human behaviour and its true complexities therefore is a good method to use to look at the more diverse and complicated areas of psychology.
   • This type of analysis is often very time consuming because there is a vast amount of written or recorded data and therefore can be very expensive to analyse due to the amount of time needed.
   • Qualitative data is hard to analyse because the data is often lengthy and hard to transfer into numerical form, as it is harder to detect patterns and draw conclusions therefore extra caution has to be taken to ensure results are valid.
   • Analysing Qualitative Data
     • In order to analyse qualitative data it needs to be put into categories to summarise the information. There are two different types of analysis content analysis or thematic analysis.
       • Thematic Analysis
         • This form of analysis is where categories are developed as the results are being analysed. The qualitative data is kept qualitative and not converted into numbers. This form of analysis is of purely textual/written data and usually involves making a written transcript and examining it in great detail for recurrent themes and meanings. Evidence for each of these is gathered from the text so that they can be discussed in turn.
       • Content Analysis
         • This is where categories are set up before the results are analysed, it takes qualitative data and converts it into a quantitative format. Text can be subjected to content analysis, where categories are set up so that frequency of occurrence of particular kinds of statement can be counted and presented using descriptive statistics. The process is normally set up as a tally chart and every time the category is stated or observed then it is added to the tally.