Mixed-Design ANOVA: Assessing the Effectiveness of Acupuncture and Massaging Programs in Treating Back Pain

Assessing the Effectiveness of Acupuncture and Massaging Programs in Treating Back Pain: Mixed-Design ANOVA

Chronic back pain has become a serious health concern for health professionals in the U.S. It is estimated that 8 in every 10 Americans will suffer some form of back pain at some time in future. For this reason, researchers are increasingly studying the causes, prevention strategies and effective treatment approaches for addressing back-related problems. This text presents the basic proposal for a research study seeking to assess the effectiveness of two commonly-used strategies for treating back pain -- acupuncture and massaging.

Assessing the Effectiveness of Acupuncture and Massaging Programs in Treating Back Pain

Chronic back pain has become a serious problem for the health fraternity in America -- the American Chiropractic Association estimates that approximately 31 million Americans experience chronic back pain at any given time, and that over 80% of adults are poised to experience some form of backache at some point in their lives (ACA, 2015). Currently, back pain stands as the leading cause of disability in the country, with the economy losing over 450 billion dollars as a direct result of the same every year (ACA, 2015). It is prudent, therefore, that health researchers focus on devising effective ways for preventing and treating chronic backache. The proposed study focuses on the treatment aspect of back pain. Acupuncture and massage are the two leading treatment modalities for back pain -- the study is geared at finding out which of the two approaches is more effective in addressing back-related problems over time. The research question guiding the study is stated as:

"Which of the two treatment approaches is more effective in treating back-related problems over time?"

We are intent on determining whether there are any significant interactions between the time taken in treatment and the treatment approach selected.

The corresponding null and alternative hypotheses are:

H0: µ A= µB -- there is no difference in the means of the two factors

HA: µ A? µB -- there is a significant difference in the means of the two factors

H0: µ A1= µA2= µA3 - there is no difference in the means of factor A (acupuncture)

HA: µ A1? µA2? µA3 -- there is a significant difference in at least two of the means of factor A H0: µ B1= µB2= µB3 -- there is no difference in the means of factor B (massage)

HA: µ B1? µB2? µB3 -- there is a significant difference between at least two of the factor B means

H0: C12 = 0 -- there is no interaction between factors 1 and 2 (time and treatment approach)

HA: CAB?0 -- there is a significant interaction between factors 1 and 2.

The hypotheses will be tested using the mixed-design ANOVA test. This test is deemed appropriate for the study because we are seeking to identify the mean differences between groups that have been split on two different factors -- time and treatment modality, where the former is a between-subjects variable and the latter a within-subjects variable (Lane, n.d.). If both independent variables were within-subjects factors (composed of unrelated, independent groups), we would have opted for a two-way repeated measures ANOVA; questions with one within-subjects factor and one between-subjects factor, however, lend themselves more effectively to the mixed design ANOVA test (Sukal, 2013).

Methods

The study will be conducted at the Cleveland Center for Spine Health in New York, which is owned and run by the researcher's family. 30 participants will be selected to take part in the study. To be eligible to participate, a patient will be required to be between 35 and 40 years of age, and should have attended regular sessions at the center for the 2 months immediately preceding the study. The 2-month eligibility requirement is a means of ensuring that potential participants and their...

A memo will be displayed at all the busy areas of the clinic, explaining to potential participants how they, and the clinic as a whole, stand to benefit from taking part in the study. Those willing to participate and satisfying the eligibility requirements will be required to register with the director, and to specify the treatment modality that they would love to take part in. Eligible participants will also be required to specify the number of massage and acupuncture sessions (if any) they have had in the three weeks immediately preceding the study. The registration window will remain open for a period of two weeks, after which it will be closed. 15 of those with the highest number of sessions in the acupuncture program, and 15 of those with the highest number in the massage program will then be selected to take part in the acupuncture and massage programs respectively. They will be notified of the same, and provided with details of how the study will be conducted via email. This systematic sampling procedure is geared at increasing objectivity and minimizing the risk of bias. The study will run for a period of 8 weeks from the 24th of October, 2015. Participants' pain levels will be measured before the beginning of the program, four weeks into the program, and upon completion of the program.
Procedures

In this case 'back pain' is the dependent variable, whereas 'time' and treatment approach' are the independent variables; however, whereas the former, time, is a within-subject variable, the latter (treatment approach) is a between-subject variable (Clark-Cutter, 2004). A within-subjects variable implies that the participants in the different stages or levels are the same (Clark-Cutter, 2004). Time is our within-subjects variable, and we will categorize it into three time points -- before the program, four weeks into the program, and eight weeks into the program. This would make it a categorical, ordinal variable. The other independent variable, treatment approach, will be defined in terms of the therapeutic approach that a participant will take part in -- it will be measured as a categorical, nominal variable, with two categories -- acupuncture program (treatment 1) and massage program (treatment 2).

We will define the dependent variable, back pain, in terms of how much a person hurts and suffers as a result of their back problems, that is how much their daily activities, movements, work, and play are affected by their back problems. The McGill pain questionnaire, a 13-item questionnaire, which requires patients to describe the intensity and quality of the pain that they are experiencing as i) mild, ii) discomforting, iii) distressing, iv) horrible, and v) excruciating, based on how much they are hurting and how much their daily activities have been affected, will be used to measure participants' pain levels before and after treatment. We will attach numerical values ranging from -2, -1, 1, 2, and 3 to the 5 responses, and the intensity of pain will be arrived at by summing the numerical values from all the 13 questions. This would make the variable a continuous, interval variable.

Results

The mixed-design ANOVA will be used to test the hypotheses stated earlier on. The pain levels for the 15 participants of the massage program and the 15 participants of the acupuncture program before the program (at time period t1), four weeks into the program (t2) and upon completion (t3) will be entered into the SPSS program, and the ANOVA test run to help the researcher determine whether there any mean differences in the pain levels of participants in the two groups, and whether there is an interaction between 'time' and 'treatment approach' on back pain. The test is appropriate because the dependent variable is a continuous, interval variable; the within-subjects factor is a categorical variable consisting of at three related categories; and the between-subjects factor is a categorical variable with two independent groups. The mixed-design ANOVA requires that the i) dependent variable be a continuous variable, ii) the within-subjects variable be a categorical variable with at least two related categories, and iii) the between-subjects variable be a categorical variable with at least two independent categories (Sukal, 2013). The study variables thus satisfy the basic conditions for a mixed-design ANOVA, which implies that the test can be used effectively.

A confidence level of .95 (p