MIDIRS Essence > March 2008 > Midwifery News
Undertaking research - starting at the beginning
Originally posted on Mar 2008
Why do it at all?
‘If you always do what you always did, you will always get what you always got’
The word research means to look again and define new knowledge. This paper offers advice for anyone undertaking research as part of their academic studies. Whether you are just setting out on your midwifery career or are undertaking a post-registration course, the framework for undertaking research is broadly the same – the difference is the level of analytical thought used in assessing the information and its relevance to practice.
What, why and how?
The first part of undertaking research is the need to develop the research question. This is rather a chicken and egg situation as to really get to the nub of what question you want to investigate you need to look at the existing evidence. It is also important that you choose a topic that interests you, you are more likely to be enthusiastic if you are exploring an issue you are passionate or even vaguely interested in than something you are not bothered about. Once you think you have something that fits the bill, you need to search the literature to see if, or how many, people have got there before you.
The following section interrupts this story to give you a few tips to help you search the published literature. If you are not sure how to begin, you might find it helpful to ask your librarian. They are used to searching the databases and can point you in the right direction. It is very helpful if you write down the form of your search, the search terms you have used and the databases you have searched for the literature. You will find a clear example of this just by reading a well constructed literature review – then all you need to do is to follow that example (eg Kunjappy-Clifton 2007)
Once you think you have grasped how you are going to start the process, the next important step is to define your topic area and identify as many connecting topics as you can at the outset – you can discard these later if you find they are not relevant, but try to keep to a reasonable limit of search terms. For example, if you were interested in Down’s syndrome, as well as looking this up as the main focus, you would probably also be interested in antenatal screening tests and chromosomal abnormalities. Depending on what aspect of Down’s syndrome you were looking into, it might also be of interest to search on pregnancy, as this would be the point for diagnosis, but also neonatal outcomes and the experiences of parents and childhood development.
Having identified the range of search terms, the next step is to consider the timeframe within which you feel the research is appropriate and relevant to your study. This is quite an important issue as there is perhaps a misunderstanding about appropriate use of research papers that were published over 10 years ago. Where there has been use of experimental work or a major scientific event, sometimes the very first papers that describe this are of use and interest when you are exploring the background for your current enquiry. However, where you are looking at evidence in support of clinical care, then, while some older papers might give you a background, it will be the most recent papers that are of use in understanding and underpinning current clinical care, especially with regard to policy and practice. A good example of this is taken from my own research where a study was undertaken in the 1950s about an experiment involving ‘early’ discharge home from hospital (Theobald 1959). This paper is of interest when discussing the context of hospital postnatal care and its history, but if your research focus was about aspects concerning outcomes related to the current provision of care, the appropriate reference would be one that offers insight into more recent events (eg Redshaw et al 2007).
The other very important aspect of the age of your references is to note the changes that might have taken place in the provision of care and in overall societal values and population characteristics. When referring to the research study undertaken in the 1950’s, this was undertaken in the UK and it was postnatal care in the UK that I was interested in. It is important to consider the location of where the research took place when deciding whether this is relevant to your own study. Taking postnatal care as an example again, where the care provision in the UK is quite different from that in many other developed countries, depending on the focus for your study, you might decide to limit the search to the UK only. Alternatively, being aware of these differences, you might choose to widen your search and compare your findings on a more international basis. When you think you have obtained most of the literature relevant to your field of study, just think once more about whether there are any aspects that have arisen from the literature search that you had not considered and whether these would be appropriate to add, or selectively discard, and make a note of your reasoning, as you will need to refer to this when you are writing up the results of your literature search.
The next steps
Having obtained the papers that are of interest to you, you will then need to refine your research focus/question; this might also be affected by the type of research method you are going to use, but just trying to clarify exactly what question you are hoping to answer is a good move at this point. Having done that, it is a good idea to search the literature again to ensure you have a reasonable range of papers; these might be for or against the basis of your research enquiry but it will be clearer to you now that you have a definite focus.
By identifying the existing literature, you can then either decide to repeat a study that has already been done but may be applied to a different population or circumstance, or, if you find very little that has been published, you can bravely go where no one has gone before – more about that tactic later.
Having found your topic
By this time you should have quite a good idea of your subject area and this should lead you to explore how you wish to obtain data that will meet your research requirements. There are two main approaches to the collection of data: quantitative and qualitative. Quantitative data (always referred to in the plural) are measurable data. The research approaches to this mainly take the form of experimental studies such as randomised controlled trials (RCTs), or descriptive studies such as surveys, all of which involve the collection of details, facts; clearly identifiable quantitative data. Taking this approach to a research study would usually involve the creation of a hypothesis or research question.
An example of this, within the context of an RCT, would be the study by Tessier et al (1998), which concluded that skin-to-skin contact from kangaroo mother care will build up a positive perception in the mothers and a state of readiness to detect and respond to their infant's cues.
Qualitative data are linked more to people’s experiences, thoughts, feelings and attitudes. These data are usually collected via one to one interviews, focus groups or observations of people’s activities. The research question is usually based on an enquiry rather than proving or disproving a matter of fact.
An example of this would be the study by Munday (2003), which explored women’s experiences of the postnatal period following a planned home birth.
So, you need to establish which approach you feel is appropriate for your research enquiry. This will then lead you to a much clearer framework for undertaking the work.
Doing a research study can be compared to making a cake
decide on what you want to make
find the recipe
obtain the ingredients & set out equipment
follow the recipe
bake in the oven
make any final adjustments and then serve
observe reactions of consumers.
People or paper?
If you plan to include people in your research, you will need to think through how you will identify and recruit them and how many people you will need as part of your research sample, some idea about the different involvement of people and whether they are actually taking part in the research as active participants. Where people take part in RCTs, complete a questionniare or take part in an interview and so have some direct link or contact with the research and the information obtained is generally contemporaneous or within recent memory. However, people can also be involved as a result of their past experiences related to their health status, the outcomes and the care they experienced. While this can sometimes be obtained through interviews or the completion of questionnaire based surveys, this information is most often obtained through access to medical records and information retrieval services.
Alternatively, research enquiry can take the form of a collation of data from existing published studies which are then subjected to a more rigorous process. The first step would be a formal literature review, recounting and discussing what is found and perhaps what is missing rather than a more in depth questioning of the research itself. This approach is used more for qualitative studies. This contrasts with the more systematic review, which explores, compares and critiques the validity of the statistical data presented, usually based on quantitative data from experimental studies.
Less common forms of research involve the researcher observing the behaviour or care of people and the circumstances in which this takes place. Data obtained from these might include measurable outcomes such as the number of times a midwife entered the postnatal ward in an eight hour period, to more qualitative outcomes, where the women in a postnatal bay discussed their views on the postnatal care they had received that day.
There is also action research, which is a proactive manipulation of situations with a resulting measurement of the effects of change on that situation. Such studies are often undertaken by a team of researchers, in a supported research based environment, but they do not need to be. If you are interested, find out more from your academic institution or do a literature search!
One last point on research approach
Both quantitative and qualitative approaches can be used in the same study; they are often seen as beneficial in exploring the research arena in its widest form rather than from just one narrow perspective. A very good example of this is where people are invited to take part in an RCT; while the quantitative aspects of the trial will record measurable outcomes, asking the participants about their involvement in the trial can sometimes reveal aspects that are important to those outcomes but were not identified at the outset. A good example of this is where participants report that where the intervention was to take a pill, the size of this meant that it was not easy to swallow, regardless of its potential benefit.
Planning strategy
The following bullet points summarise the issues you need to think about as you plan how to undertake your research
Research question/hypothesis - What do you want to know and how will you measure or explain the value of your findings?
Sample size and characteristics - Who is going to give you the information and how much do you need for validity
Data format - Facts and figures, words/ language, visual observations
Analysis - What will you do with the data when you have got it
Filling in the answers for each of these points will be specific to the nature of your research. I noted earlier that it is not uncommon to repeat a research study to see the effect of it on a different population. If this is the direction you decide to go in, most of the work will have been done for you with regard to the research approach and methods to use. However, if you need to, or want to undertake a totally original piece of work, this is a one-off, and you will need to think carefully about each step of what will be involved.
If you come across a problem, and possibly even if you don’t, you should seek advice from your study supervisor so that you are clear you are heading in the right direction before you put a lot of work into something that might not be able to give you the answers you are looking for – read on…
How do you get at the truth?
Bias
Establishing the potential for, or actual forms of bias is one of the essential tasks for researchers. The essential question you need to ask yourself before you begin any form of data collection is what ‘unseen’ factors might affect your findings. The following aspects of the research design might have an impact on this.
Having decided on who (or what) will be your main source for data collection you need to determine the sample size. Different research approaches will mean different sample sizes and while there appears to be a common view that bigger is always better, this is not always the case. For the collection of qualitative data, where this involves vast quantities of written text from interviews, the in-depth nature and quality of the data requires this to be obtained from a smaller number of people. The number of participants for qualitative studies may not be decided from the outset but may be led by factors that suggest there is no more information to collect – this is called saturation of the data – so the researcher stops.
For quantitative studies, this is usually determined by statistical estimates of the number of people required to be involved in order to establish that the findings were due to the effect of the experiment or intervention from the trial and not due to chance. The power of the study determines this based on the frequency of the outcomes; the rarer the outcome, the more power (more people) is needed to demonstrate this. A good example of this is the Term Breech Trial (Hannah et al 2000). Where an uncommon event is being studied, then any experimental study would require very large numbers of women to be involved.
Analysis and statistical significance involves looking for or discovering causes and associations with the data collected, the characteristics of the participants and the role of chance. If you are collecting measurable data it is always a good idea to find a statistician to help you decide what statistical tests you should use, and to help you establish the sample size based on power calculations where these are needed.
The aim of all of this preparation is that your findings should be as free from bias as you can make them and will therefore be valid and reliable. In qualitative studies this is referred to as trustworthy. In working your way through this process you need to be able to answer the following question:
Can you justify the methods used to support your analysis and the interpretation of your findings? If not, what needs changing?
Have you asked nicely?
If you are involving people or the information about people in your research you are legally obliged to obtain ethical approval before you start. You can get information about the local requirements for this from your Trust or academic institution’s Ethics Committee.
You will be required to provide a full explanation of how you plan to undertake your research in the form of a research protocol. It is this that will be approved, which means you will then need to keep to it. Ethical approval will involve consideration of how you are going to obtain consent on an individual basis from prospective participants or their parents/guardians and how you will manage the collection of personal details, the maintaining of confidentiality or anonymity and the secure storage of data containing personal information. The Ethics Committees will also want to see some consideration of how you plan to disseminate your final results, so you need to think about feedback to your participants as well as plans to publish your findings within a reasonable timeframe in a recognised publication.
Support
There are many sources of support, but you may need to find them rather than expecting them to find you. This is to some extent the nature of the research as an individual enquiry is just like individual care. Your needs with regard to the research will be quite specific and this is best addressed by seeking out someone who is prepared to offer you one-to-one support on a fairly regular basis. This should be implicit within your institution, but you should also think about other resources apart from your lecturers. These might include librarians, statisticians and people with expertise in other departments, for example psychology, social sciences and technology in the form of video & visual aids.
Apart from the literature review that focussed on your research topic, there are a range of other written resources in the form of textbooks and other published studies and articles about the research process. Then there are your peers, so share your experiences, the good and the bad and help each other.
A whole new language
The following is just a guide to avoiding getting into difficulties in the first place or getting out of them if you do!
If you don’t know what it means - look it up
Still in the dark? - ask a colleague
Not much clearer ? - ask an expert
Clear as the light of day? - repeat the process but try different words!
This underpins the information given above about seeking advice. Some of the research terminology is complex and can be quite full of research jargon so fully understandable when you know what it means, but quite incomprehensible when you don’t; but as with any foreign language – there will be an interpreter somewhere.
Speed humps: understanding the research language
Here are some tips to help you as you begin your research journey. Good luck!
Data - information sources, usually used in the plural when referring to them in the text.
Hawthorn effect – where the research process affects the behaviour of the people being studied – usually in its favour and therefore potential for bias.
Hermeneutic philosophy – an interpretation of phenomena.
Homogeneity – peas in a pod – inclusion of similar characteristics.
Paradigm – (the g is silent) a particular perspective or point of view.
Phenomena – a process, effect or concept under review.
Phenomenology – a research approach that aims to explore the ‘lived’ experiences of individuals from their own perspectives.
Triangulation – a combination of two or more research approaches or analytical techniques in one study to achieve better understanding of the research question.
Variables – characteristics being studied, either selectively included as part of the research, excluded (controlled) or accidentally uncovered.
The general rule is if the word is unfamiliar – look it up! Most research textbooks have excellent glossaries
Some resources:
MIDIRS: offer standard searches, special searches and photocopies of selected articles. Plus the MIDIRS Midwifery Digest and much more!
Cluett E, Bluff R eds (2006). Principles and Practice of Research in Midwifery. 2nd ed. Edinburgh: Churchill Livingstone. £24.99.
Rees C ed (2003). An introduction to Research for Midwives 2nd ed. Oxford: Books for midwives. £16.79.
Robson C (2002). Real world research: a resource for social scientists and practitioner-researchers. Oxford: Blackwell. £23.74.
Wickham S (2006). Appraising research into childbirth: an interactive workbook. Edinburgh: Elsevier. £22.99.
Davies MB (2007). Doing a successful research project. Basingstoke: Palgrave Macmillan. £12.99.
Proctor S, Renfrew M eds (2000). Linking research and practice in midwifery: a guide to evidence-based practice. London: Bailliere Tindall. £23.99.
References:
Hannah ME, Hannah MJ, Hewson SA et al (2000). Planned caesarean section versus planned vaginal birth for breech presentation at term: a randomised multicentre trial. Lancet. 356 (9239):1375-83.
Kunjappy-Clifton A (2007). And father came too...a study exploring the role of first time fathers during the birth process and to explore the meaning of the experience for these men. MIDIRS Midwifery Digest. 13(3):371-5.
Munday R 2003 What are women’s experiences of the postnatal period following a planned homebirth : a phenomenological study. MIDIRS Midwifery Digest.13:3 371-5.
Redshaw M, Rowe R, Hockley C et al (2007). Recorded delivery: a national survey of women's experience of maternity care in 2006. Oxford: National Perinatal Epidemiology Unit.
Tessier R, Cristo M, Velez S et al (1998). Kangaroo mother care and the bonding hypothesis. Paediatrics 102(2):e17.
Theobald G (1959). Home on the second day: the Bradford experiment. Br Med J,/i> 2(5163):1364-7.
www.midirs.org
Dr Sally Marchant | Editor MIDIRS Midwifery Digest | MIDIRS
Your Comments
Have your say! Register now and post your comments on this article here.
Sorry, there are currently no comments for this article.