This article is being published in the electronic journal, Psychology of Espionage Reports, which can be found at The Psychology of Espionage Reports is a sub-part of Dr. LeRoy A. Stone’s website, which can be found at:

Editor’s Note -

The article, Married Couple Spies, was the product of a investigation, conducted by this Journal’s Editor over half a decade ago, upon the request of some ‘interested parties’ following the uncovering of Aldridge and Maria Ames as spies. The written article being published here is almost exactly in the same form and version as it was when it was written back in the 1994-95 period of time. There was seen no need to change much of anything even though this traitorous spy database now (i.e., as of 2002) has had several more entries added to it.  Also, it is important to note that more married couple spies have not been uncovered since 1994 when the original version of this article was first written. This can be said as Aldrich and Maria Ames are the last known discovered married spy team who have been entered into the spy database.

Married Couple Spies

LeRoy A. Stone*
West Virginia University**


     Even though the Soviet Union, which formally was the major nemesis of the U.S.A., crumbled a few years ago, U.S.A. citizens, who are traitorous spies, continue to be uncovered. As an example of this, Aldrich and Rosario Ames were arrested on 21 February 1994 for their spying that originally was to assist the Soviet Union but later was to serve Russia. The writer has assembled, over a several year period of time, a U.S.A. traitorous spy database, in which all contained information has been quantified so that it may be suitable for statistical analyses. It was built from information obtained from the nonclassified or open-literature sources (i.e., newspapers, magazines, books, electronic journalistic data bases, informational requests based upon the Freedom of Information Act, etc.). As of the time of the writing of this paper, the traitorous spy database included information for 232 spies, who were caught/uncovered since the end of World War II, up to and including 1994. This database, as well as its development has been previously described in a number of publications (e.g., Stone, 1991; 1992) and presentations made at scientific meetings (e.g., Stone, 1993).

     Most of the 232 spies described in this database were unmarried and, if married, their spouses were not known to have been aware of or were involved in the spying efforts of their husbands. However, of the total number of the known spies, 26 (or 11% of the database) of them were in marriages in which both husband and wife were both jointly identified as being spies. Again, both Aldrich and Rosario Ames would be the most current examples of such a married duo of spies. Aldrich pled guilty of committing espionage and Rosario pled guilty to conspiracy to commit espionage as well as evading taxes.

                                                                                        - 2 -

     Other such familiar-name couples (along with the year in which they were caught or identified) who are rather infamous would be Ethel and Julius Rosenberg [1950], Anne and Jonathan Pollard [1985], Karl and Hanna Koecher [1984], and John and Barbara Walker [1985]. Although Barbara was never charged with any spying type offenses, it is well known that she was an active and knowing participant and assistant in many of her husband's spying activities.  Other such couples, whose names may not be as well known (or with one couple their names for many years were classified) are: Ruth and David Greenglass [1950], Myra and Jack Soble [1957], Martha Dodd-Stern and Alfred Stern [1957], Bevalyn and William Hugle [1983], Ruby Schuler-Harper and James Harper[1983], Jane Foster-Zlotovsky and George Zlotovsky [1957], Lona and Morris Cohen [1961, and who also were known in England as Helen and Peter Kroger], and Spy Six and Spy Six’s Wife [1948]. These last two are now believed by some to have been named: Mr. and Mrs. Clarence Hiskey, however this has never been officially verified. All of these named spies (i.e., 13 couples) were U.S.A. citizens at the time they were involved in espionage activities.

     In an investigatory effort to try to understand spies and their spying behaviors, a research question was posed. Are spy couples and their spying behaviors different from spies and their spying behaviors who operate in a solo fashion independent from cooperative spying efforts by an involved spouse? Answers to such a question could in fact prove to be very valuable and useful information. Such kinds of information could provide behavioral science foundations for the development of a number of varied counterintelligence strategies as well as add to the dearth of limited information, based upon actual behavioral science research, that is known regarding the psychology of spying or espionage.

     Statistical exploration of the U.S.A. traitorous spy database was carried out in a fashion that rather simply investigated the possible relationships of the behavioral/biographical variables (that are 87 in number) on which the 232 spies have been measured/scaled with the one variable that did indicate married or single marital status. Initially, the distinction of married vs. single was studied. Then the database variables were examined with respect to only those who were married to one another; i.e., more specifically, this analysis had to do with married to one another vs. being single or being simply married, but not to another spy. Lastly, analyses were accomplished which provided statistical predictions (based upon behavioral/biographical information) of whether an uncovered spy might be in a marriage with another spy vs. not married to another spy.

     A rather elementary statistical procedure known as ‘Student’s t-test’ was initially employed to examine many of the database variables regarding whether single or married spies differed appreciably (i.e., to a degree that could be regarded as being statistically significant) on each of the involved variables. For example, they did differ, as one would expect, on the gender variable. Proportionally, more male spies were single than were female spies. In fact, just about all of the female spies were married.

                                                                                      - 3 -

     The results were found that significantly differentiated (based upon use of the Student’s t-test paradigm) single vs. married spies.  These significant dfferences were multiple and are summarized in the remainder of this paragraph. Married spies were more likely to be civilians, they seldom were seen as being military (especially so in the U.S. Navy), they were older, they were more intelligent, they had served more years in defense or security employment, they were involved in spying for longer periods of time, they started their spying when older, they were higher status positioned employees, there was less homosexuality history in their backgrounds, they were known to exhibit more pronounced foreign preferences, their mental health was more in the favorable direction, they had fewer drug abuse histories, they were more likely to have Jewish backgrounds, quite expectedly - they had more accomplice associations, they started their spying more years ago, they were more prone to be influenced into spying because of ideology rather than by greed, and they were less involved with substances (undefined) abuse. Of course the single vs. married groups differed in other matters but not to a statistical significance level degree (i.e., beyond what is known as the 0.05 level, which implies that the differences observed were greater than one would expect less that five times in a hundred, if only chance were operating) than was observed with the variables mentioned earlier above in this paragraph.

     The differences that were found between the single and simply married vs. spies married to one another classifications were also examined using the same t-test statistical tool. Differences that were regarded as being statistically significant (determined following the very same logic a mentioned in the last sentence of the preceding paragraph) are summarized in the remainder of the present paragraph. Spies married to one another (when compared to spies who are single or are simply married to non-spies) are seen as more prone to be civilians rather than in the military, possess more education, are more intelligent, less likely to be motivated by greed (and hence are more motivated based on ideology), have less years of national defense/security employment, were involved in their spying activities for much longer periods of time, exhibited greater foreign preferences, tended to have Jewish backgrounds, their security clearance levels were lower or sometimes even non-existent, they had fewer money problems, they tended to be later written about or described to a lesser degree in U.S. Government publications (in the open literature), more of their spying activities took place more in the continental U.S.rather than in overseas locations, and they started their spying more years ago. As mentioned before, database variables, other than those described immediately above did indicate differences between the two studied married groups, however the extent of the differences did not reach levels that could be regarded as being statistically significant.

                                                                                       - 4 -

     In a further attempt to examine the relationships that a particular collection of variables had with married spies and with spies married to one another, a statistical relationship/prediction methodology known as multiple correlation or regression was employed. Multiple correlation results in a coefficient that expresses the degree of relationship (where 0.0 indicates no relationship and 1.0 indicates perfect relationship) between a dependent variable and a pool of independent or predictor variables. Such a statistical methodology can be used to ascertain the complex relationships that may exist between variables as well as to ascertain in the overall relationships which independent (i.e., the predictor) variables are significantly related to the dependent (i.e., the predicted) variable.

     When the multiple correlation denoting the 10 security clearance adjudication variables (i.e., based upon the logic of the adjudication criteria set forth in the publication: Director of Central Intelligence Directive, 1/14) with the dichotomous classification of unmarried spies vs. married spies, a highly significant relationship degree of association ( i.e., R = 0.65) was observed. When the statistical significance of the independent or predictor variables were examined in the overall relationship model, it was found that the following described security clearance adjudication variables made significant (i.e., at the 0.05 significance level) contributions to predicting whether the studied traitorous spies were married or not. Married spies were shown to exhibit greater indications of holding more foreign preferences than were seen with unmarred spies. Married spies were seen to have histories of more financial problems than were seen with unmarried spies. Less sexual misconduct was observed with married spies than with unmarried spies. More alcohol abuse, but less drug abuse was observed with married spies. Less prior criminal conduct was seen with married spies than was seen with the unmarried spies. Less mental/emotional disturbance was also seen with the married spies as compared to the unmarried spies. It does not appear that married spies differed significantly from unmarried spies with respect with extent of having of foreign relationships, having made more record falsifications and having committed more security violations or the like.

    When the same 10 security clearance adjudication variables were studied with respect to their relationship with the dichotomous classification of single and married spies vs. spies married to one another, another highly significant association was observed. When the statistical significance of the independent or predictor variables was examined in the overall relationship model, it was found that the following described security clearance adjudication variables made significant contributions to predicting whether the studied traitorous spies were single or simply married vs. whether they were married to one another. Spies married to one another were shown to be more likely to have shown foreign preferences than were those spies who were single or simply married. Also, the spies who were married to one another were observed to have a more pronounced history of alcohol abuse than were those spies who were single or were simply married, but not to one another.

                                                                                       - 5-

     The predictive accuracy obtained when using the statistical methodologies (which are known by statisticians as discriminant functions) described in the immediately preceding two paragraphs, was seen as being well beyond what would be expected, simply based upon chance. For example, using dichotomously scaled information pertaining to the 10 security clearance adjudication variables it was possible to predict from the couple hundred studied spies, those who were married vs. those who were single with relatively low prediction error rates. When predicting simply whether the spies were married or not, a 93% prediction accuracy rate was observed with respect to the correct prediction of those spies who were simply noted to be married (which also includes those spies married to one another). When predicting whether the spies were married to one another an 85% prediction accuracy rate was observed with respect to correct prediction of those spies who were married to one another. These prediction rates are regarded as being statistically significant beyond what is known as the 0.05 level, which implies that they are not likely to be based upon some type of chance arrangement.

     The research findings reported in the preceding paragraphs would seem to carry or contribute some rather important type understandings regarding traitorous U.S.A. citizen spies who have operated since the end of World War II (i.e., back to the end of 1945). For one thing, the overall collection of such spies who are known to have operated during the past almost 50 years would appear to be composed of identified subgroups that bear characteristics that make one rather different from another. The reported research has shown that the smaller group of spies who are married to one another do appear to be quite different, in a large number of differing respect that are those spies who are single or who are married to spouses who are not spies. Even the greater subgroup, which consists of all those spies who are married (to either one another or to spouses who are not spies), is seen as representing a rather real different population (in a multitude of respects) than are those spies who are single. The smaller subgroup of spies who are married to one another appears to represent a more ‘professional’ and more ‘successful’ group than those who are single or those who may be married but not to another spy. Such should be not at all surprising as a married couple spy team consists of no less than two spies whereas the more singular spies represent the spying efforts of only person. In the case of traitorous spies, two appears to be better than one!

     One understanding that seems forthcoming from these findings is that the so-called married couple spies not only seem to be a more successful and professional subgroup that can be found in the known U.S.A. traitorous spies group, many of them also seem to be more well-known. Some of this fame (or infamy) most likely may be due to the fact that a few of these married couple spies have been discussed in the popular literature as having successfully stolen many of our Country’s most important secrets. As a result, the married couple spies may perhaps be regarded as a more dangerous (i.e., to our Country) subvariety of traitorous spies.

- 6 -


Stone, L. A. (1991, Oct.). I spy a myth. Security Management, 35, (No. 10), 26-32.
Stone, L. A. (1992). A principal components analysis of 59 variables descriptive of uncovered spies. In D. Sands & J. Ellis
   (Eds.) Proceedings of the 34th Annual Conference of the Military Testing Association, Vol. II, (pp. 576-581). San
   Diego, CA: Navy Personnel Research and Development Center.
Stone, L. A. (1993). Intelligence level for U.S. citizens identified as spies. Read at the American Psychological Association
   meeting, Toronto, Canada, August, 1993.


   *Dr. Stone’s mailing address is: P.O. Box 395, Harpers Ferry, WV 25425-0395
** He holds the academic rank of Clinical Assistant Professor in WVU’s School of Medicine’s Department of Behavioral
     Medicine and Psychiatry.

Click Here to Return to the Index Page
for Psychology of Espionage Reports

Click Here to Go To the Index Page of
Dr. Stone's Web Site