PSYC 4850: What I Have Learned

I was nervous for this class as “Advanced Methodology” sounds terrifying, but I just needed a 4000 level psychology class to graduate. Little did I know how much I would enjoy it. I found that this class was not scary at all and that it was actually one of my favorite classes to go to! Who knew I would be interested in methods, not me. I thought that advanced methodology would entail solely math and statistics, but I found that there are many routes you can take when looking at methods. It was interesting to see what path everyone took as we all have different interests. I think that it is very valuable that we were able to use our strengths and interests to study and present our findings. I found that we were able to bring our strategies and unique intelligence into our work, where this may not happen in a lecture based class. I enjoyed that there were no exams, as I become very stressed during exams. There was still a lot of work for this class, but it was work that I could work on at my own pace and that I enjoyed. Over the course of this semester I have found that I can teach myself a lot on my own. I have noticed that self-taught classes are very helpful, in my opinion, because you have to force yourself to do things, whereas in lecture you can zone out and not listen if you choose.

For the chosen topic, I personally have always struggled with reading graphs and understanding the information that is portrayed in these displays. This is why I took on this topic as I thought it would be a challenge to help me learn all of the different ways of displaying and reading data. I can confidently say that I have learnt a lot about all of the different methods and their disadvantages and advantages. I also learned the rules and guidelines in creating these graphs, which will come in useful in my psychology degree and maybe in a career. I was not interested in this topic before but now I think there are so many cool graphs that can tell us a lot. Graphs are used not only in scholarly articles, but also in the news, textbooks, newspapers, marketing, etc., which shows how important they are to understand.

I learnt a lot from reading all your blogs and I thank you for teaching me so much about your various topics. I also thank you all for taking the time to read my blogs and posting awesome comments! It was very nice that no one was judgmental and everyone was so accepting with all blogs and talks. Because of this, presenting got easier and easier as the semester progressed. It has been a great class and I really enjoyed learning from all of you whether it from a blog entry or a talk.

 

Synthesis of Displaying Data

The display of data is present in many psychological articles. It is helpful to present conclusions and organize, summarize, simplify, or transform data (Verdinelli, 2013). The two most common data displays in articles are graphs and tables (Scott & Mazhindu, 2005). Both are used to summarize and present the data in an easy to read way (Scott & Mazhindu, 2005). Combining everything I have discussed regarding displays, I would theorize that displaying data is useful to provide clear presentation of statistical data, which is extremely important in articles. Schmid & Schmid (1979) state that agencies spend a large amount of money in gathering statistics, but far too frequently this effort is wasted if there are no graphs or tables shown. This is wasted because attention is not paid as much to the analysis when display presentations are not included. Charts and graphs can translate data into attractive, concise, and understood illustrative forms (Schmid & Schmid, 1979). These forms possess qualities and values that are lacking in the text. I will go over some of these values below:

1.) Well-designed charts are more effective in creating readers interest and in appealing their attention than other types of presentations.

2.) Charts that display statistics represent an important form of visual communication because they are clear, economical, and precise for conveying a message.

3.) Visual displays are superior to words or figures because they are more clearly grasped and easily remembered!

4.) Charts and graphs provide a more complete and better-balanced understanding that could not be derived from text.

5.) Use of displays saves time à large amount of data into a small form.

6.) Charts and graphs can bring out hidden facts and relationships, which can stimulate analytical thinking and further investigation.

The reason I chose this topic was because anyone interested in research should be knowledgeable in the principles of constructing graphs and charts. One may not actually be drawing these charts, but will need to plan them and know when to use and not to use certain graphic methods. I am not saying that articles should solely have graphs, I am saying that articles should have text and graphs to work together in order to display and share information more powerfully.

To end off my topic of displaying data, I want to share with you a fun type of display. This is the band chart, also known as the stratum chart due to its layers. It is a set of line charts where data is aggregated or disaggregated so that the distance and area between two lines represents the amount of a certain variable (Schmid & Schmid, 1979). Each band of information is placed above the other (Beckford, 1998). Strata charts are useful to show a visual representation of features in a process and display the importance of the various parts of the process (Beckford, 1998). The example below is a multiple-strata surface chart:

References

Beckford, J. (1998). Quality: A critical introduction. London: Routledge.

Schmid, C.F., & Schmid, S.E. (1979). Handbook of graphic presentation. New York, NY: Ronald Press Publication.

Scott. I., & Mazhindu, D. (2005). Displaying data. Statistics for Health Care Professionals.

Verdinelli, S.V. (2013). Data display in qualitative research. International Journal of Qualitative Methods, 12, 359-381.

Displaying Data: Presenting and Interpreting Results

When a study is finished and has been analyzed, the important question to ask is: what information should be presented? It is not practical to display every single small detail, however it is important to not leave out any critical information regarding the study. The researcher should then figure out which meta-analytic model should appear. A meta-analysis is a research methodology and statistical technique that aims to quantitatively integrate the results of a study (Meca & Martinez, 2010). The next questions to consider are: what are the best techniques for displaying the meta-analytic results? How do the meta-analytic results reflect the theoretical analysis? Visual displays can assist in the interpretation of meta-analytic results (Reis & Judd, 2000). Displaying data is essential to analyzing the data (Scott & Mazhindu, 2005). The display type is linked with the type of data because some displays are better for qualitative versus quantitative, for example. Displays help establish how the data is distributed and to find any trends. To visually examine study outcomes enhances the potential for finding differences in the data (Reis & Judd, 2000). After choosing the visual display that will be used, a researcher should look into whether the synthesis uncovered important areas in the literature that deserves future research.

Funnel plots are like scatterplots, but are used for looking at sample sizes versus effect sizes from studies (Reis & Judd, 2000). They simultaneously display a sample statistic and the sample size (Rakow, Wright, Spiegelhalter, & Bull, 2015). The name “funnel plot” came from the idea that the estimated effect increases as the size of the study increases. This is like a funnel because effect sizes from smaller studies will scatter more widely at the bottom of the graph and the larger studies will narrow, resembling an inverted funnel (Higgins & Green, 2011).

The effect estimates are plotted on the horizontal scale, and the measure of study size on the vertical axis (Higgins & Green, 2011). This is the opposite of conventional graphs, when normally the outcome is plotted on the vertical axis and the covariate is plotted on the horizontal axis. If there were a publication bias, a funnel plot would reveal fewer entries in the effect size portion of the graph (Reis & Judd, 2000). If there were a bias because smaller studies that had statistically significant effects were not published, for example, this would show an asymmetrical funnel plot.

When used in psychology or medicine, funnel plots are helpful to communicate risk; however they are less effective as a decision aid for individual patient’s treatment decisions. Funnel plots are useful because when the more pronounced the asymmetry is on a plot, we can tell that it is more likely that a bias is significant. A limitation that funnel plots have is that sometimes the effect estimates and standard errors are naturally correlated, which can produce spurious asymmetry in the funnel plot. Below is an example of a funnel plot:

References

Higgins, J., & Green, S. (2011). Cochrane handbook for systematic reviews of interventions. The Cochrane Collaboration. Retrieved from http://www.chochrane-handbook.org

Meca. S.J., & Martinez, F.M. (2010). Meta-analysis in psychological research. International Journal of Psychological Research, 3, 2011-2084.

Rakow, T., Wright, R.J., Spiegelhalter, D.J., & Bull. C. (2015). The pros and cons of funnel plots as an aid to risk communication and patient decision making. British Journal of Psychiatry, 106, 327-348.

Reis, H.T., & Judd, C.M. (2000). Handbook of research methods in social and personality psychology. New York: Cambridge University Press.

Scott. I., & Mazhindu, D. (2005). Displaying data. Statistics for Health Care Professionals.

Displaying Data: Line Graphs, Scatterplots, Boxplots, and Histograms

Continuing my topic regarding data displays and presentation of results, this week I will discuss other types of displays. When I perform research I notice that line graphs, scatterplots, and boxplots are also used in articles. These displays efficiently present a large amount of information using little space (Owl Purdue). I will briefly talk about each below:

Line Graphs:

Line graphs display correlations between quantitative variables, with each point representing the mean score of the dependent variable. Bar graphs and line graphs are very similar in that you could simply place bars that reach the points and it would be very similar. The reason we have both is because one would use a bar graph when the variable plotted on the x-axis is categorical and use a line graph when it is quantitative (Price, 2012). Below is an example of a line graph:

Scatterplots:

Each point in a scatterplot represents an individual rather than the mean for numerous individuals. There are no lines connecting the points in a scatterplot, however sometimes a regression line is used to show the best fits of the points. A linear regression line is not always used but it can be if the researcher wants to use one. A question that came to mind when looking at the individual points was: what if two or more individuals fall on the exact same point? One could resolve this by offsetting the points slightly along the x-axis, display the number of individuals in parentheses next to the point, or making the point larger or darker in proportion to other dots (Price, 2012). Below is an example of a scatterplot:

Boxplots:

Boxplots are useful for comparing distributions and identifying outliers. They use boxes and lines to create a representation of a distribution. The box displays the main scores, the dots (if used) show where the actual data are, and the lines show the scores at the middle and edges of the distributions. The box covers the middle half of the data, the line inside the box is at the median, and anything outside of these are outliers. Below is an example of a boxplot:

Histograms:

Histograms are the last common type of data representation display that I will discuss. They are graphical displays that show what proportions fall into which categories. The categories are displayed as bars and must be adjacent. Histograms are similar to bar graphs except that they are used for continuous data, as opposed to qualitative. The area of each bar is proportional to the frequency and each have equal width.

Pie charts are rarely seen in journal articles (Peers, 2006). This is because it is more difficult to compare angles and sections of a pie chart than heights and lengths of bars in a bar graph; therefore it is not recommended. Each kind of display is useful in its own way, and this is why there are the different types. The one that I find the most helpful in looking at clusters and trends are scatterplots with a regression line. The dots are helpful in seeing where there are clusters of individuals and the regression line helps to display any trend. Scatter plots give a good visual of all the variables and are a useful summary of the set of data.

References

Owl Purdue Online Writing Lab. (n.d.) APA tables and figures. Retrieved from https://owl.english.purdue.edu/owl/resource/560/19/

Peers, I. (2006). Statistical analysis for education and psychology researchers. London: The Falmer Press.

Price, P.C. (2012). Psychology research methods: Core skills and concepts. Retrieved from http://2012books.lardbucket.org/pdfs/psychology-research-methods-core-skills-and-concepts.pdf

Displaying Data: Tables

Continuing from last week I will resume talking about different ways of displaying research data. As I explained in my last blog, many journal articles will have the data displayed in another way other than text/writing. I discussed bar graphs and how they are simple yet very useful and well known across many disciplines, including psychology. While doing my own research I have noticed that many journal articles have the data displayed in tables. Cooper, Hedges, and Valentine (2009) explain that using tables to present findings is the most common method to display results. These tables can be very simple or very complicated, depending on if they are displaying all of the results or specific areas. I think that tables are beneficial in articles because space is limited in journals and readers find it challenging to process large amounts of detailed data, so therefore tables are helpful to help explain the results (Peers, 2006). Below is an example of a table one would find in a psychological article:

There are specific guidelines for tables being published in articles within journals. This makes it easier not only for the public readers but also for the journal editors and reviewers. APA-style tables do not have any vertical lines, only horizontal (Myers & Hansen, 2011). Tables should appear in the methods or results section of a published article, and not in the body (Myers & Hansen, 2011). The tables provided must be necessary and not put in to take up space (Owl Purdue). All tables should be numbered (Table 1, Table 2, etc.) with every column having a heading (Owl Purdue).

Some articles will use a multitude of different types of displays in a single study. Tables provide a summary of the important aspects of the data, and graphs convey the information visually, therefore it can be beneficial to use both (Peers, 2006). The article below is an example that shows the use of both a table and a bar graph:

Data can be displayed in tables whether there is one response variable or several. When a data distribution for one response variable is displayed it is called a univariate distribution (Peers, 2006). These statistics show features such as the mean and the standard deviation (Peers, 2006). An example of a univariate analysis would be a researcher wanting to investigate differences in final exam performance, among different groups. This is a univariate analysis because the research question relates to whether there are differences between groups regarding a single response variable. A bivariate distribution is when two variables are shown, which is commonly displayed in a scatterplot, which I will discuss in one of my next blogs. Bivariate distributions show the degree of relationship between two variables (Peers, 2006). Multivariate is the analysis of multiple response variables, which can be related or different from each other (Peers, 2006). Going back to the example of the final exam results, multivariate analysis would look at the performance on final exams as one response variable, the salary of employment could be a second variable, and the number of previous exams completed could be an independent variable. A kind of multivariate analysis is repeated measures analysis where for example one could look at the performance at the end of one year, two, and three (Peers, 2006). All of these types of data analysis can be displayed in articles and are common in demonstrating results.

References

Cooper, H, Hedges, L.V., & Valentine, J, C. (2009). The handbook of research synthesis and meta-analysis. New York: Russell Sage Foundation.

Myers, A. & Hansen, C. (2011). Experimental psychology. Wadsworth: Cengage Learning.

Owl Purdue Online Writing Lab. (n.d.) APA tables and figures. Retrieved from https://owl.english.purdue.edu/owl/resource/560/19/

Peers, I. (2006). Statistical analysis for education and psychology researchers. London: The Falmer Press.

Displaying Data: The Beginning

In my next four blogs I am going to share the different ways of displaying data. There are several methods to present data, however I will only discuss four common types over the next four weeks and go into depth with them, with this blog including one.

A single statistic only shares a portion of any results, whereas a well-designed statistical display of data helps to understand relationships taking place. When there are a large number of results to report, it is more clear and efficient to do so with a graph of some sort (Price, 2015). We all learn in our school years how to interpret and read different types of data sets, its part of the curriculum. Even just learning how to understand a simple bar graph comes in useful in the future. We see data sets being displayed everywhere, whether it is in the media (newspapers, television, online), scholarly journals, books, the workplace, schools, etc. Displaying data sets in a visual way rather than written is helpful to visualize any trends or patterns. Additionally, it is very difficult to share all the data in words, whereas a visual can encompass all of it as a whole.

When information of any kind is being posted, we must ask the question: How should the data be presented? The answer to this question depends on where the data is being posted, because this will change the way in which it is shared. For example, if data results were being posted in a scholarly journal, we would see many complex graphs because of the audience and complexity of the research. Whereas, if it was in a newspaper in which the general population is reading, then the display of data would be different and simpler so that everyone can understand it. Moreover, depending on the type of data there are useful ways to display certain data, as I will discuss.

I will begin by discussing the bar graph, which is one of the techniques used to present data in a visual form so that the viewer is able to see any trends or patterns (Statistics Canada, 2013). This sounds and looks simple but there are many factors of a bar graph. Bar graphs are used constantly in articles and we see them almost everywhere. Bar graphs can be either vertical, horizontal or stacked (Statistics Canada, 2013). The bars portray different values, with an X and Y-axes to represent the scale.

This is an example of a vertical bar graph:

This is an example of a vertical bar graph with two series of data:

This is an example of a horizontal bar graph: 

This is an example of a horizontal bar graph with two series of data:

This is an example of a stacked bar graph:

The following is an example of a bar graph including error bars to represent the standard errors. A standard error is the standard deviation of the distribution (Price, 2015). These error bars are used because then one can see whether a difference is statistically significant or not (Price, 2015).

A disadvantage of bar graphs is that researchers can choose a specific graph in order to manipulate data, making individuals misinterpret the results. An example of this is when researchers create the X or Y-axes as increasing very rapidly or slowly, so that we think something is drastic. When looking at bar graphs it is important to look at the axes and take the variables into consideration before making a judgment. The most common and best use of bar graphs is dealing with frequency distribution and time-series statistics (Statistics Canada, 2013). Most people can read and understand bar graphs, and thus they are a widespread way of displaying data.

There are also many requirements when presenting statistics in bar graphs. The American Psychological Association (APA) has guidelines, which includes the idea that the reader should be able to understand the results based solely on the graph and its caption, without having to refer to the text for explanation (Price, 2015). There are also technical guidelines for graphs including rules regarding layout, axis labels, legends, and captions. For layout guidelines the graph should be somewhat wider than it is tall, the independent variable should be plotted on the x-axis and the dependent variable on the y-axis, and values should increase from left to right on the x-axis and from bottom to top on the y-axis (Price, 2015). The axis labels have to include the units of measurement if they are not already in the caption, the labels should be parallel to the axis, and legends should be placed within the boundaries of the graph (Price, 2015). Captions are supposed to briefly describe it, explain any abbreviations, and captions in an APA manuscript are to be presented on a separate page at the end.

Though bar graphs are simple, they are important in displaying data. It is also important to understand the guidelines for publishing articles and results. Bar graphs are used in many psychological research papers and are perhaps the most common method of depicting data (Newman & Scholl, 2012). Although as Newman & Scholl (2012) state, one must not give in to the tendency to depict bar graphs as asymmetric, as many bar graphs are depicted in a certain way to gain a certain objective. I think that bar graphs are very useful and helpful in understanding data and observing results, however as with anything one must look into the details. Many do not understand the misinterpretations and therefore believe what they see, which is not a good thing.

References

Bar graphs. (2013). Statistics Canada. Retrieved from http://www.statcan.gc.ca/edu/power-pouvoir/ch9/bargraph-diagrammeabarres/5214818-eng.htm

Newman, G.E. & Scholl, B.J. (2012). Bar graphs depicting averages are perceptually misinterpreted: The within-the-bar bias. Psychonomic Bulletin & Review, 19, 601-607.

Price, P.C. (2015). Expressing your results. Research methods in psychology: Core concepts and skills. Retrieved from http://catalog.flatworldknowledge.com/bookhub/18?e=price_1.0-ch12_s03

Developmental Research

Understanding the varying circumstances in which we develop is an important concern in psychology (Purta, 2014). Developmental research is focused on the progressive changes that occur as an individual develops. This type of research is significant in psychology because it provides information about individual diversity. Many factors that take place in childhood can affect one’s future and can largely shape their life. There are two different methods of developmental research with both contributing to qualitative findings. The first type is cross-sectional studies, which compare different samples of people at one time to find contrasting versus similar occurrences. Normally in cross-sectional research, the individuals will share characteristics such as socioeconomic status, educational background, and ethnicity. Thus, the researcher can see what characteristics exist, so it does not determine cause and effect relationships between variables. The other type of developmental research is longitudinal, which looks at the same individual or group over an extended period of time (Cherry, 2015). In longitudinal research investigators will collect data before they begin the study and then may gather data repeatedly throughout (Cherry, 2015). Longitudinal research can last anywhere from a week to several decades (Cherry, 2015). Depending on what the researcher is looking for they can decide whether to perform cross-sectional or longitudinal research methods.

Longitudinal research is helpful to look at how certain variables impact people. Jones, Joshua, and Lawrence (2011) performed a study, which is a helpful example of a developmental study, where an intervention is performed and the impact is observed. The experiment was two years long, where they observed the impact of a universal integrated school-based intervention in social-emotional learning and literacy development on children’s social-emotional, behavioural, and academic functioning (Jones et al., 2011). There were 1184 children from 18 elementary schools, of who were randomly selected. Results found that the intervention showed improvements across several domains. This is an example of a longitudinal study because they started with the same children and followed them throughout the two years and then viewed ending results. A longitudinal study fits the goal of this study, whereas a cross-sectional study may not work as well for this specific study. Cross-sectional research provides a snapshot of a population at one point in time and does not find how something is changing over time, but can compare observational characteristics at a specific time. An example of when a developmental cross-sectional study is useful in psychology would be when researchers want to find out how children’s ability to perform a memory task varies according to age.

Many people are critical of developmental research because they prefer the idea of systematic mechanisms and processes (Kloep & Hendry, 2014). However, I think that developmental research is helpful because in longitudinal studies one can see the large picture of someone’s life, instead of just performing a study in the present. Longitudinal studies are specifically beneficial because researchers can look at changes of the same people over time and through this they can observe lifespan issues. Cross-sectional studies are valuable as well because they can be gathered quickly and researchers can develop highly informative studies without the time associated with longitudinal work. Both types of observational data are helpful in understanding the development of certain people or groups over time or at specific times.

References

Cherry. K. (2015). What is longitudinal research? Retrieved from http://psychology.about.com/od/lindex/g/longitudinal.htm

Jones, S.M., Brown, J.L., & Lawrence, A.J. (2011). Two-year impact of a universal school-based social-emotional and literacy intervention: An experiment in translational developmental research. Child Development, 82, 533-554.

Kloep, M., & Hendry, L.B. (2014). Some ideas on the emerging future of developmental research. Journal of Adolescence, 37, 1541-1545.

Purta, J.T. (2014). Nested societal contexts and the application of developmental research. Journal of Applied Developmental Psychology, 35, 478-480.

Replication in Research: Why is it Important?

Replication involves repeating a study by using the same methods, but using different subjects and experimenters. Conducting replication in research is vital because repeating experiments checks for reliability. True replication is impossible, however it is attempted to be as precise as possible (Lash, 2014). Well-established and accepted research is replicated to correct for one-time occurrences. The original researchers should provide an operational definition, which explains how they performed the research (Dewey, 2007). This helps the individuals performing the replication because they must know how the original researcher accomplished measurements.

If there are failed results after replication, this does not always mean that the research is poor and that the researcher lied. This may be because there is a difference in the way the research was performed during the replication. Other reasons that might influence the different outcome of a replication experiment could be due to minor details such as the weather, location, time of day, or the instrumentation (Dewey, 2007). These differences should be looked into to find out the reason why it happened; which is the main reason why replications are performed. If a researcher is hesitant in having replication studies performed on their work, this may be suspicious. I found a few cases where researchers were requested to share their data and they refused. This makes one wonder why, and sometimes the individuals that attempted to perform the replication will publish the refusal as a way to publicize it.

By conducting replications we are reducing biases because biased research will provide favored ideas, and replications can find this out. Replication can help detect fraud, but when replication results fail it does not always mean there has been fraud, it can spring new ideas for research (Dewey, 2007). Replication is an incentive for researchers to create honest research and reduce biases, because they may get “caught” (Dewey, 2007). When originally creating a study, researchers should make it so that they are providing enough details to permit replication. Replication can be most difficult in the field of psychology. When trying to repeat a scientific experiment in the hard science fields such as chemistry, biology, and physics, there are concrete variables and results; therefore replication can be consistent in these fields. However, in psychology many argue that replication is not that simple and is hard to replicate due to biases, susceptibilities, individual’s feelings, and priming.

I think that replication studies are important to test for validity and to find any errors regarding reliability. Many psychological studies would not be easy to replicate because of the biases that can occur with individuals. Additionally, I believe that any extension and replication of a study is beneficial to the results and idea being tested.

Questions to think about: If a study cannot be replicated, what value does it have? Can you be confident in its data? Should we not trust studies that are not replicated? Is one study enough to claim evidence? Can we trust replicated studies more?

References

Dewey, R. (2007) The importance of replication. Retrieved from http://www.intropsych.com/ch01_psychology_and_science/importance_of_replication.html

Lash, T.L. (2014). Research through replication. Paediatric and Perinatal Epidemiology, 29, 82-83.

Observational Research

Observational research is also known as field research and is a type of non-experimental investigation where the researcher observes behaviour without intervention (Hammer, du Prel, & Blettner, 2009). Many laboratory settings create artificial environments in which one or more variables are manipulated for the experiment. In contrast, the majority of observational research looks at the observed participants in the natural setting. Studying behaviours that occur naturally in non-altered environments where variables are not manipulated is the main difference between clinical trials and observation.

The main drawback to using this approach versus an experiment is that you do not have control over everything, so one cannot manipulate a variable while holding another constant. Therefore it is harder to isolate the causal effects of a particular variable. So instead, individuals generally try to look for this kind of thing happening naturally in order to approximate an experiment without having to set one up. The advantage with observational research is that it is easier and also feasible in many cases where experiments are not. Researchers have to make sure they include enough observations because any individual observation could be a rare occurrence, so for example if the variable is very low it is hard to be confident of the results.

Selection bias is a very present weakness in observational research (Hammer et al., 2009). Selection bias arises when the observed population is not a random selection (Hammer et al., 2009). Randomization is difficult in observational research because the individuals are in a specific place at a specific time and investigators cannot alter this. Depending on where the observation is taking place, culture, religion, and class, for example, may result in the outcome being bias towards certain groups. This tends to generalize the results by having culture and social status, for example, as being similar among those being observed (Hammer et al., 2009). The solution to this problem is to gather enough data and observations so that one can control for these factors.

I would argue that observational data is more qualitative than quantitative. Aagaard and Matthiesen (2015) discuss the idea that qualitative research has only been related to language-based research such as interviews or recordings. However, they argue that observation is qualitative. Both Aagaard and Matthiesen are qualitative educational psychological researchers and disagree with the definitions of qualitative research, due to its emphasis on language. They argue that qualitative psychology research is mostly limited to using interview research, however observation is very qualitative research as well, and should be used more. I agree with this point because in interviews the participant has all the power in their speech and can be bias. Observation can remove the bias of the language used in interviews by both researcher and interviewee.

Although observational research faces many limitations and can be vulnerable to researchers’ biases, it still makes an important contribution to psychological results and knowledge (Hammer et al., 2009). Observation is a qualitative type of research, which is needed, however research groups should attempt to decrease issues such as the selection bias. There are some things that cannot be studied in a laboratory setting, so we have to use observational data as the next best option.

References

Aagaard, J. & Matthiesen, N. (2015). Methods of materiality: Participant observation and qualitative research in psychology. Qualitative Research in Psychology.

Hammer, G.P., du Prel, J.B., & Blettner, M. (2009). Avoiding bias in observational studies. Deutsches Arzteblatt International, 106, 664-668.

Research Ethics in Pediatric Psychology

While researching the topic of ethics I came across something that I thought was really interesting: research involving minors. Conducting pediatric research is far more complicated than research on adults due to consent and assent issues. Research on children also poses important challenges with regard to vulnerability and possible conflicts of interest (Fernandez, 2008). The main question that comes to mind when I think of pediatric research is, do minors have their own say and how much input do they have? Researchers legally cannot rely on the child’s full consent, but depending on their age they will get a child’s assent (Fernandez, 2008). Assent is the concept of a minor providing agreement to their participation in a study where full consent is not possible due to age (Fernandez, 2008). Legal authority by parents or guardians is made to provide permission for their child to take part in a research study (Field & Behrman, 2004).

Something that strikes me as being inconsistent is that researchers cannot rely on the child’s consent if he or she is a minor in their country, however the minor is legally able to get a job and drive a car. Investigators, when appropriate, try to involve children in discussion about the research, but at the end of the day they do not have the right to give the go ahead (Field & Behrman, 20014). These inconsistencies are shown in the ongoing debate about the boundaries of the ethics of parental permission. This debate is especially prominent when it involves adolescents. A topic that is often discussed is at what age investigators should start getting the child’s assent. I think that it should be earlier than the age of 18 because these individuals are making many other decisions that are more significant and potentially more dangerous. The studies that are taking place are not considered dangerous due to other existing ethical policies that are in place. The pediatric ethics board facilitates the quality of their interactions with the children and the methodological quality of their research to administer a safe environment for the children. If the facilitator goes over the benefits and risks in detail and the child gives written and oral consent they should not need permission from a parent after a certain age; this age could be 16 for example. Having the age as 18 is inconsistent with the other decisions they make in their lives.

With the consent being vested in parents or guardians leaves the question: why are they assumed to have the best interests of the child if the child is old enough to know what is going on? Why is the child seen as not knowing his or her own best interests? If a child is very small and cannot make his or her own decisions then there should be informed consent from the parent or guardian. When a child is older the majority of stress of the consent should be placed on the child, instead of the parent having most of the say. When the child is an infant and unable to make decisions then the parent can have the say.

References

Fernandez, C. (2008). Ethical issues in health research in children. Paediatrics & Child Health, 13, 707-712.

Field, M.J., & Behrman, R.E. (2004). Ethical conduct of clinical research involving children. Washington (DC): National Academies Press (US).