Sample Letter to a Graduate Student, Statistical Consulting Center

May 1, 2002

Rowdy Raider
1234 Every Day Drive
Dayton OH 45678

Dear Rowdy:

This letter is a report of the meeting that you, Dr. Noit Ahl, and I had on April 29, 2002, about your master's thesis in nursing. For your thesis, you plan to conduct a descriptive, correlational study to determine the relationship between perceived satisfaction with educational opportunity and organizational commitment, overall work satisfaction, and intent to remain in the organization among staff nurses at a regional medical center. Nurses whose primary job is to provide direct patient care and have clinical or non-managerial responsibilities will be included in the study.

You expect between 43 and 119 nurses to participate.

You plan to defend your proposal in either December 2002, or January 2003. Data collection will begin as soon as possible after the defense. You anticipate the data collection will be completed by March 2003. You will be graduating in June 2003.

You will be using four instruments in your study, Perception of Hospital Educational Opportunity Scale (Kirsch, 1990), Organizational Commitment Scale (Monday, Steers, & Porter, 1979), Job Description Index (Smith, Kendall, & Hulling, 1969), and Intent to Remain Scale (Price & Mueller, 1981).

The Perception of Hospital Educational Opportunity Scale has 16 (7 negatively worded and 9 positively worded) statements regarding nurses' perceptions of hospital sponsored staff development activities. Respondents rate the statements using a 7-point Likert scale from 1 = "Strongly Disagree" to 7 = "Strongly Agree." After reverse scoring the negatively worded statements, responses are summed and divided by 16 to provide a mean score. Higher scores indicate a more positive perception of educational opportunity.

The Organizational Commitment Scale has 9 positively worded statements about the feelings a nurse may have about the employing organization. Respondents rate the statements using a 7-point Likert scale from 1 = "Strongly Disagree" to 7 = "Strongly Agree." A mean score is calculated as it was for the Perception of Hospital Educational Opportunity Scale. Higher scores indicate a more positive level of organizational commitment.

Additionally, you constructed a demographic questionnaire to determine age, rank, gender, highest nursing degree held, and length of time as a nurse.

Frequencies and frequency percentages will be calculated on all variables in your study. Additional descriptive statistics such as the mean, standard deviation, median, mode, minimum, maximum, and range will be calculated for age, length of time as a nurse, and scores for all four instruments.

Research Question #1: What is the perceived level of satisfaction of staff nurses with educational opportunities at a regional medical center?

Descriptive statistics on the scores for the Perception of Hospital Educational Opportunity Scale will be used to answer this question.

Research Question #2: What is the level of organizational commitment of staff nurses at a regional medical center?

Descriptive statistics on the scores for the Organizational Commitment Scale will be used to answer this question.

Research Question #3: What is the relationship between the perceived satisfaction with educational opportunity and organizational commitment, overall work satisfaction, intent to remain, and selected demographic variables?

Multiple regression analysis will be conducted to answer this question. Indicator variables for gender, educational level, and rank will be entered into the model as independent variables. (The actual number of indicators needed for educational level and rank will be determined by how many different levels are represented in the sample.) Additionally, scores for organizational commitment, overall work satisfaction, and intent to remain will be entered into the model as independent variables. Perceived satisfaction with educational opportunity as measured by the Perception of Hospital Educational Opportunity Scale will be entered into the model as the dependent variable.

Since educational level and rank may be correlated, the Variance Inflation Factor (VIF) indicating multicollinearity will be checked to ensure the assumptions necessary for using multiple regression analysis are met. If the assumption is violated, then only one of those two factors will be included in the model.

In calculating necessary sample size, I assumed one variable would be needed for gender, one for educational level, and three for rank, for a total of 8 independent variables in the analysis. The power analysis calculations assume that you are going to use all 8 variables in the regression model and that you want to know if any one variable is significant given that all of the other 7 are already in the model. The overall alpha level was set at 0.05. I set different levels of overall R-squares or total amount of variance in the data explained by the variables already in the model. Also, I assumed that you would want to detect an increase of 5% in explained variance for the variable to be significant.

To achieve 80% power with overall R-square of 0.50, you need complete data on 80 participants.

Preliminary power calculations were performed based on the formula in an article by Sande Milton (see reference below.) Final power calculations were performed using PASS version 6.0.

Also during our meeting, I pointed out your options for data entry. If you would like us to enter your data into a spreadsheet for you, our charge for this service is $25.00 per hour. (The estimated time to enter your data is 2 1/2 minutes per questionnaire.) Your other option is to enter the data yourself. We accept data in a variety of forms (Excel, Qedit, Word and others.) In either case, please consult with Mary Alspaugh in our office prior to your data entry..

The only other charge is $30.00 per hour for data management. Data management is SAS programming to format the data to make it useful for analysis as well as to label the data for easier reading.

If you have any further questions or if you feel I have not understood your goals, please let me know. Also, I need to be kept informed of any changes that you may make to your study. I enjoyed meeting you and look forward to working with you on this project. Good luck with your work!

Sincerely,

Stacie E. Taylor
Statistical Consultant

SCC #2001035

cc: Dr. Noit Ahl

Reference:
Milton, S., 1986. A Sample Size Formula for Multiple Regression Studies, Public Opinion Quarterly (50), 112-118.