There are many theories that psychologists suggest to explain the difference in prevalence rates of Major Depressive Disorder in men and women.  First, there are biological theories - that is, arguments that there are legitimate physiological differences between men and women that account for the higher rate of depression in women.  There are also psychosocial theories that argue that women have different types of interactions with people than men do - and there are also demographic differences.  For example, Nolen-Hoeksema argues that women ruminate more (worry more) about relationships; Brown & Harris (1978) argue that women experience more social stressors which make women more vulnerable to depression.

However, there is also the argument that the prevalence of depression is equal among men and women but that women report and seek help for depression more than men do. This is known as the artifact hypothesis. This hypothesis also argues that there is a tendency on the part of clinical psychologists to overdiagnose depression in women. 

Padesky and Hammen (1977) carried out a study with university students.  Male and female university students were given a series of scenarios, demonstrating various intensity of symptoms linked to clinical depression.  For each scenario, they were asked to rate the degree to which the person needed to seek treatment.  The researchers found no significant difference between the two genders.  Both said that they would seek out treatment at a comparable level of depressive symptoms.

However, this study was a hypothetical scenario, not real life.  Although the study had high internal validity, it did not have high ecological validity.  In order to attempt to evaluate the validity of the artifact hypothesis under naturalistic conditions, Amenson and Lewinsohn carried out their study. 

The sample was made up of 998 participants who were recruited through an announcement mailed to 20.000 residents of Eugene and Springfield, Oregon. The names were randomly selected from the county voter registration list. Participants were told that they would be part of a study of "the understanding of psychological health and its relationship to what people do, think and feel." 

There were two times that the participants were assessed for depressive symptoms.  First, they were asked to fill out a 938-item questionnaire and mail it back to the researchers.  The second time was about 8 - 9 months later when they were interviewed at the clinic. The interview was a 2-hour semi-structured interview.  The interviewers were blind to the questionnaire data. Part of the questionnaire was the Center for Epidemiological Studies Depression Scale (CES-D), which measures the intensity of depressive symptoms.

To test the artifact hypothesis, men and women were divided into high, medium, and low symptom level groups based on their CES-D scores. Each group had a similar mean score. When asked about how they labeled their own behaviour and whether they had sought help, there were no significant differences between the genders in any of the three groups.

In addition, they compared the CES-D scores based on self-reported symptoms with the clinical diagnosis from the two-hour interview. The analysis found that men and women with equal reported symptom levels were equally likely to be diagnosed as depressed, regardless of whether the interviewer was male or female. Self-labelling and clinical diagnosis were in agreement for 81% of the female and 92% of the male participants.

This study supports the argument that there are actual differences in the prevalence of depression in males and females and that is not the result of reporting or clinical biases.

Women diagnosed as depressed less frequently labeled themselves as depressed. This means that studies that rely on self-reported data may actually underrepresent the level of depression in women.

The study was done under naturalistic conditions, where participants discussed their own health.  This study is high in ecological validity.

The study was done in the US - a culture that is highly individualistic.  It would be difficult to generalize the findings to other cultures where self-reporting of depressive symptoms may be more stigmatizing. The study may lack cross-cultural validity.

In addition, the study is over 30 years old.  It is possible that education and the integration of psychology into high school programs means that people are more aware of depression and are more willing to seek treatment.  The study may lack temporal validity.

The researchers also found no differences when dividing participants by income, educational level or employment status. This also challenges the psychosocial theory to explain gender differences in prevalence.

Padesky, C.A., Hammen, C.L. Sex differences in depressive symptom expression and help-seeking among college students. Sex Roles 7, 309–320 (1981). https://doi.org/10.1007/BF00287545