Skip to Main Content

Research Process Guide

Step 9: Analyzing Data

Once you collect the data, you need to analyze the data. Depending on your methodology and your research questions, you will determine how you will analyze the data.

What is Data Analysis?

For most researchers, data analysis involves a continuous review of the data. Analysis for both quantitative and qualitative (numerical and non-numerical) data requires the researcher to repeatedly revisit the data while examining (Kumar, 2015):

  1. The relationship between data and abstract concepts.
  2. The relationship between description and interpretation.
  3. The data through inductive and deductive reasoning.


Regardless of your methodology, these are the 4 steps in the data analysis process:

  1. Describe the data clearly.
  2. Identify what is typical and atypical among the data.
  3. Uncover relationships and other patterns within the data.
  4. Answer research questions or test hypotheses.


Quantitative data analysis

The first thing that you want to do is discuss a step-by-step procedure for the analysis process. For example, Gall et al. (2006) outlined steps used to review the pretest and posttest design with matching participants in the experimental and control groups:

  1. Administer measures of dependent variables to research participants.
  2. Assign participants to matched pairs on the basis of their scores based on step 1.
  3. Randomly assign one member of each group to the experimental groups and the other to the control group.
  4. Administer the experimental “treatment” to group.
  5. Administer the measures of dependent variable to the experimental and control groups.
  6. Compare the performance of experimental and control groups on posttest using tests of statistical significance.

Then you want to tell the reader about the kinds of statistical tests that will be implemented on the dataset (Creswell & Creswell, 2018):

  1. Report descriptive statistics including frequencies (i.e., how many male, female, non-binary participants?), means (i.e., what is the mean age?), and standard deviation values for the primary outcome measures. Standard deviation is formally designed as the average distance to scores away from the mean.
  2. Indicate inferential statistics test used to examine the hypothesis of your study. For experimental design with categorical variables you might use t-tests or univariate analysis of variance (ANOVA), analysis of covariance (ANCOVA), or multivariate analysis of variance (MANOVA). There are several tests mentioned below categorized based on your measurement.
Tip: Remember that all Kean University faculty and students have access to the statistical analysis software SPSS. It would be important to note that your statistical analysis will take place entirely in that SPSS format. You can administer the test below, based on your research goals and objectives.


Kinds of statistical analysis:

Using software like SPSS, you can conduct statistical tests to examine your hypothesis and research questions (Bryman & Cramer, 2009; Ong & Puteh, 2017; Kumar, 2015):

  1. Nominal measurements:
    • Frequency distribution
    • Proportions/percentage values
    • Chi square
  2. Ordinal measurements:
    • Median
    • Percentile rank
    • Spearman rank order
    • Correlation
    • Mann-Whitney test
  3. Interval measurements:
    • Mean
    • Standard deviation
    • Pearson's Product-moment
    • Correlation
    • Inferential procedures (T-tests, ANOVA)
  4. Ratio measurements:
    • Geometric mean
    • Percentage variance
    • Inferential procedures (T-test, ANOVA)


Qualitative Data Analysis

Qualitative data analysis, unlike quantitative, does not require hypothesis testing but rather deals with non-numerical data, in the form of words. Qualitative data is inductive, therefore at its root, it is about creating a theory or understanding through analysis and interpretation (Miles et al., 2018; Bogdan & Biklen, 2007). Traditional qualitative researchers have historically completed their data analysis by hand. There is something about getting your hands dirty with the data. That said, there are other options for qualitative data transcription and analysis. Also, there are many methods of coding. Below is a review of the general process of coding techniques.

All qualitative data analysis involves the same four essential steps:

  1. Raw data management - "data cleaning"
  2. Data coding cycle I - "chunking," "coding"
  3. Data coding cycle II - "coding," "clustering"
  4. Data interpretation - "telling the story," "making sense of the data for others"


Transcription - Raw data management

Qualitative data usually involves transcribing interviews or focus group data that has been previously recorded with participant consent.

Tip: Transcription is very time consuming and it is important that it is exact. During the time when in-person research could not be conducted due to Covid-19, many platforms have been used to record and transcribe data from interviews and focus groups. Zoom platform has the option to automatically transcribe the data in real-time. You will have to clean the data a bit by reviewing transcripts and making sure that what participants said was accurately transcribed (Think: subtitles). A fail-safe measure to make sure your data is valid is to ask participants to review the transcription themselves, in order to make sure that the perspectives and experiences that they shared in answering the questions are accurate.
Tip: It is also important to note that traditionally qualitative data collection methods required researchers to transcribe and code “by hand,” meaning typing up the transcripts and then coding on hard copies of the data. Today, there are many different qualitative data analysis software programs to choose from. These programs assist in transcribing, organizing, coding, and analyzing data. Some of the most well-regarded software programs are:
 
  1.  MAXQDA
  2.  NVivo
  3. ATLAS.ti
  4. QDA Miner
  5. Quirkos
  6. Dedoose
  7. Taguette

For the record, at this time, Kean University students and faculty do not have access to any qualitative data analysis software. However, there are opportunities for free trials or occasionally a deeply discounted licensing fee for students for the programs listed above.


Data Coding - Cycle I

There are several steps in the first cycle of coding. The first thing you need to do is to immerse yourself in the transcript data. Read the data. Read it again. Then, read the data again. Do this several times, and as you do so, you will start to get a sense of the data as a whole. Start annotating in the margins, “chunking” data into categories that make sense to you. This step is your very first preliminary pass at coding. As you do the “chunking,” read over the chunks and see if you start to identify patterns or contradictions. Some really excellent guides and resources for you as you begin your coding process are:

The Coding Manual for Qualitative Researchers, 4th edition - Johnny Saldana (2021)

            Qualitative Data Analysis: A Methods Sourcebook, 4th edition - Matthew Miles, Michael Huberman, & Johnny Saldana (2020) 

Both of these books are held by Kean University's Library.
 

Tip: Code book is a manual of all the codes that you use. A code book identifies and defines code names and explains the protocol for what data is included and what data is not included. You will begin with 25-35 codes. As you move through the cycles of analysis, your codes will be combined into categories and then themes.
Tip: If you are using a data analysis software tool, you will be able to do each set of coding cycles within the program. Essentially, the steps and processes are the same as if you are coding by hand. The process for each software cycle will vary depending on the program.

Data coding - Cycle II

During your second cycle (and third, if need be) of coding, you start clustering chunks of data that have similarities. As you are doing this, you are reading over the chunks of data, refining your code book, and narrowing down the scope of each code.  You will go through 2 or 3 cycles of narrowing down codes, grouping them together, and winnowing down the data. You will most likely move from 25-30 codes to grouping them together in clusters to develop themes. These themes are the core of your data analysis. You will end up with 5-7 central themes that tell the story across the data (Saldaña, 2021).

Kinds of Coding

As you work your way through the data analysis, you will be going through three different kinds of coding as you progress (Miles et al., 2018; Bogdan & Biklen, 2007; Creswell & Creswell, 2018):

  1. Open Coding - assigning a word or phrase that accurately describes the data chunk. You do open coding line by line in the transcriptions of all interview data. This is the first coding step.
  2. Axial Coding - is the process of looking for categories across data sets. Takes place after open coding. More in the second or third cycle of coding. Remember: you cannot categorize something as a theme unless it cuts across data sets.
  3. Cluster Coding - taking chunks of data that share similarities and review and code in several cycles. Reduce codes by removing redundancies. Here you are refining your code book to develop themes across data.

References

Bogdan, R., & Biklen, S. K. (2007). Qualitative research for education. (5th ed.). Allyn & Bacon.

Bryman, A., & Cramer, D. (2009). Quantitative data analysis with SPSS 14, 15 & 16: A guide for social scientists. Routledge/Taylor & Francis Group.

Creswell, J. W., & Creswell, J. D. (2018). Research design: Qualitative, quantitative, and mixed methods approaches. Sage.

Gall, M. D., Borg, W. R., & Gall, J. P. (2006). The methods of quantitative and qualitative research in education sciences and psychology. (A. R. Nasr, M. Abolghasemi, K. H. Bagheri, M. J. Pakseresht, Z. Khosravi, M. Shahani Yeilagh, Trans.). (2nd ed.). Samt Publications.

Kumar, S. (2015). IRS introduction to research in special and inclusive education. [PowerPoint slides 4, 5, 37, 38, 39,43]. Informační systém Masarykovy univerzity. https://is.muni.cz/el/1441/podzim2015/SP_IRS/

Miles, M. B., Huberman, A. M., & Saldaña, J. (2018). Qualitative data analysis: A methods sourcebook. Sage.

Ong, M. H. A., & Puteh, F. (2017). Quantitative data analysis: Choosing between SPSS, PLS, and AMOS in social science research. International Interdisciplinary Journal of Scientific Research, 3(1), 14-25.

Saldaña, J. (2021). The coding manual for qualitative researchers (4th ed.). Sage.