Technical

The Organization Diagnostic Instrument (ODI) Analysis

Although the data collection component of the ODI is as simple and streamlined as possible, there is nothing simple about the level of analysis provided to our clients.

Once the data has been collected, we import results into SPSS for analysis. SPSS (Statistical Package for the Social Sciences) is one of the most advanced statistic tools available on the market, and is the core of all our data analysis activities.

While it may vary slightly from client to client, the most common steps in our analytical framework are as follows:

Definition of Sample Characteristics: Prior to beginning any analytical work, we define the critical sample frame needed for all subsequent analytical work. This includes the identification of the sample size, the sample completion rate and response rates and confidence intervals based on a standard 95% margin of error .

Basic Statistical Analysis: Our initial analysis begins with a ‘basic’ exploration of the data, primarily looking at central tendency, variance and data distribution. This is done on a question-by-question level, and looks at the mean, median, mode, standard deviation, standard error, skewness, and kurtosis levels. This fundamentally tests each question for accuracy, error, precision and bias.

Scale Condensation: After basic exploratory analysis is conducted on the data set, we condense all of the frequency data into a single ‘mean score’ for each question. This creates the broader range of responses usually needed to make deeper statistical analysis more effective and comprehensible. In order to do this, each of the Likert responses is assigned a mathematical value:

Strongly Agree = 100 Agree = 75 Neutral = 50 Disagree = 25 Strongly Disagree = 0

All values for each question are summed up and divided by the number of respondents to each question, thereby providing a mean score per question that will range from 0 to 100. These scores not only better reflect the totality of the responses, but also have the added advantage of factoring out Don’t Know / No opinion and No Response answers that can impact later analysis.

Correlations: Once we are sure that all of the standard statistical tests are within tolerances, we look at the level of correlation between all of the questions. This is done using standard Pearson Tests. This gives us a good understanding of how response patterns to different questions may be related and suggests which of the more advanced statistical techniques used later will be the most appropriate to use on your data set.

Multivariate Analysis: During this most advanced step of the process, we conduct a variety of multivariate analytical techniques in order to identify the level of predictive influence the questions and dimensions have on each other. These techniques most commonly include multiple regression analysis, factor analysis, and principle component analysis, but could include other forms of analysis depending on the data set’s size and cohesion. The exact techniques selected at this phase of analysis would be completely dependent on earlier sample characteristics, since many common statistical techniques require at least 20 respondents to produce reliable test output.

If you have fewer than 20 employees, or we get fewer than 20 responses to the survey, we can still run the ODI for you, but we are severely limited in what multivariate techniques we can use. We would then be limited to reporting frequencies and mean scores for each question, dimension and layer. While this will still give you highly useful information, we would not be able to identify the causal relationships among the various dimensions, which is one of the core benefits of the ODI as a diagnostic tool.

The ODI Report

After completing all of the analysis, we will provide you with a report in Microsoft Excel that provides you with the following information:

The Overall ODI Scan Results

The ODI Scan Results are the cornerstone of our diagnostic analysis. They are a simplified picture of the ODI results that enables rapid and easy assessment of core strengths and weaknesses occurring within your organization. The Scan gives the mean score by layer and dimension, along with a color code. An example ODI Scan can be seen below.

ODI Scan Results

Red Zones: Those dimensions with a mean score between 0 and 25. These areas represent an acute weakness in the organization that is adversely affecting business operations and the ability of the organization to achieve its goals, to a significant degree.

Yellow Zones: Those dimensions with a mean score between 26 and 50. These areas represent a modest weakness within the organization that is likely affecting business operations and the ability of the organization to achieve its goals, to a moderate degree.

Green Zones: Those dimensions with a mean score between 51 and 75. These areas represent normal healthy functionality in the organization.

Blue Zones: Those dimensions with a mean score between 76 and 100. These areas represent deep strength in the organization and are likely a source of strategic competitive advantage.

Question Details

In addition to the ODI Scan Results, all data is provided on a question-by-question basis. Question level data includes:

1) Exploratory results for each question. This includes all measures of central tendency and distribution we use in the initial assessment of the data. Generally it will include mean, median, standard deviation, standard error, skewness and kurtosis values. 2) Frequency counts for each question. This is the number and percentage of survey respondents who selected each of Strongly Agree, Agree, Neutral, Disagree and Strongly Disagree for all questions. 3) Net positive and Net negative scores for each question. This is the combined number and percentage of those who Strongly Agreed or Agreed combined for Net positive and the combined number and percentage of those who Strongly Disagreed and Disagreed for Net negative. This makes for much easier comparisons across questions. 4) Mean Score for each question. The condensed data score for all questions, ranging from 0 to 100.

Multivariate Results

Where the data set is large enough to support multivariate techniques, a final result table with all of the derived coefficients will be provided for each multivariate technique employed. Unless specifically requested, only final output will be provided since including all of the intermediary steps used in gaining the final results for most multivariate techniques would be cumbersome and uninformative.

Qualitative Results Mapped by Dimension

The final question posed by the ODI is an open-ended question designed to collect qualitative data that can be mapped against the various statistical dimensions. Employees are encouraged to provide a detailed written response. Each comment collected is read individually and a primary code is assigned to the central theme of the comment based on one of the ODI dimensions: salary, operational management, vision, human resource systems, etc. Once all the data is coded, all responses with a specific code will be presented in a separate tab broken out by central theme. This is an incredibly powerful tool, as it helps to paint a narrative picture of what the numerical score in the ODI Scan Results only quantifies mathematically. With both the numerical and the narrative results of the survey, you will be well equipped to develop evidence-based strategies and action plans to improve the weaker areas of your organization.

Helping to Build Strong & Healthy Organizations Worldwide

Contact Us today to get your diagnostic started.